SIGNLD in the news
Interviews, podcasts, and articles featuring the SIGNLD team.

Chris Rafter on IBM's Making Data Simple
Episode 423: Drowning in data? Governance, third-party data, and Decision Intelligence.
Hosted by Al Martin, WW VP Technical Sales, IBM · November 4, 2026 · 51 min
Listen to the full episode
In Chris's words
“People think they have to have all their data ready and cleaned on day one, and then a starter's gun fires and it's go time. SIGNLD is designed to meet you where you are.”
“Do you have seven departments or eight departments? Because when I read QuickBooks, it said you had eight, but you just told me you have seven, so which is it?”
On how the Knowledge Graph asks questions back.
Transcript
Transcript lightly edited for readability.
00:00Al MartinOkay, podcast listeners, Al Martin here again. Week after week, we do this every week, and you believe it. The title for today, or at least how I'm going to characterize it, is the data revolution: why companies have more information than ever, yet they still struggle to make better decisions. Chris, who I'll introduce in a second, is an interesting guest because he sits at the intersection of business strategy, data analytics, AI, and revenue growth. His background at Inzata will give us a chance to explore both the technology and some of the practical business challenges. I'm here with Chris Rafter, who is the Chief AI Officer for SIGNLD, which is part of Inzata, and he'll explain this in a second, where he serves as the Chief Operating Officer. Chris, thank you so much for being here. I appreciate you, man.
00:54Chris RafterThank you, Al. Appreciate you having me on.
00:56Al MartinLet's start with you. Tell us about yourself. How did you get into the technology business in the first place? And don't leave out any of the good stuff. Tell us where you started and what landed you here.
01:07Chris RafterSure. I'm Chris Rafter. I'm COO and a co-founder at Inzata Analytics, and we recently launched a product called SIGNLD, which has turned into its own brand, so I'm also Chief AI Officer there. It's a platform designed to turn a company's scattered data into a Knowledge Graph that AI can actually reason over, and use that context to help them get better value out of SIGNLD's AI or any other AI tools or agents they might be using.
01:34Al MartinFantastic. But you're like a serial entrepreneur, aren't you?
01:37Chris RafterA couple, yeah. I started my career in the Fortune 500 and financial services, actually, and got pulled into technology during the last big bubble we had, the internet bubble. I've been in technology ever since. I started getting involved in data about 25 years ago, building data marts, data warehouses, and things like that. I have a lot of experience across multiple industries, big companies and small and medium-sized companies, as well as a lot of public sector work with state governments, state agencies, and large and small K-12 school districts. So I've seen what's different and what's common across data environments in multiple verticals and industries.
02:11Al MartinIf I did my research correctly, you've been involved in technology, operations, revenue, product, marketing. A pretty broad collection of responsibilities. Which one have you enjoyed most?
02:24Chris RafterThe customer-facing stuff. I'm usually involved on the pre-sale side, but also in post-sales and delivery, and my favorite moment is a few weeks after onboarding, when the customer is getting results, they're happy, and they come back and tell us what it meant for them. Maybe they're able to better predict graduation rates in a K-12 school environment and help students out, or they're able to save money in a large government organization, or get people better services more accurately and on a more timely basis. Those are the moments that are most memorable for me.
02:55Al MartinThen conversely, which one gives you the biggest headache?
02:58Chris RafterThere's a lot of education that goes with the job, and a lot of times you're having to explain things. Sometimes people aren't ready to hear the truth about things like data quality and master data management, and how important that is to virtually any result or objective they're going to set. Those are moments where you have to slow down, meet them where they are, find out what they do believe, and then try to bridge that and move forward from there.
03:26Al MartinWhat's your brand? Are you a data guy? Is that what you say? Are you an operations guy? It could be anything.
03:31Chris RafterI'd probably lean more toward data, product management, innovation. What I really love is talking to people, listening to them talk about their challenges, the things they find difficult and the things they wish they could do differently, then taking that back, turning it over in my head, talking about it with my team, and coming up with solutions to those problems that they'll appreciate or get value out of.
03:55Al MartinThat's perfect. That's your answer. As long as that's your brand, that's your brand. But let me ask you this. If you could go back and give your 25-year-old self one piece of career advice, what would it be?
04:08Chris RafterOh, gosh. It probably would have been to jump into entrepreneurship sooner. I spent most of my career in corporate roles as an employee, and not moving into entrepreneurship earlier is something I definitely regret. I would have done maybe five or ten years, but doing fourteen was probably a little too long. I would have just taken the leap much earlier.
04:33Al MartinWhy is that? What's the attraction? Money? More control, more broad experience?
04:39Chris RafterWell, I wish it were money. No one's going hungry. It's definitely the autonomy and the stimulation you get from doing lots of different things. One of the popular sayings in a startup is that there's no such thing as "it's not my job." Whatever has to get done has to get done, and I've always been that way. I've always been objective-focused. I'd much rather be given a goal and told, here, go attack that goal, here's some budget to do it, and I'll check back with you in six months. I think a lot of people are like this. They appreciate the autonomy, like to think on their own and chart their own path, and then participate in the benefits when they achieve something.
05:17Al MartinVery good. That's a good answer. Let's talk about Inzata. Tell us the story behind Inzata. What problem did you see in the marketplace that made you think, man, there's got to be a better way to do this?
05:26Chris RafterSure. It's still a problem today, and it may actually get worse. The problem we set out to solve was that way more data was being collected and generated than was being put to practical use to support decision making. It's been growing at an exponential rate, and companies' capacity to process that data, get it into a usable state, into a data warehouse, a data lake, a data mart, dashboards, and ultimately in front of the people who need it to make decisions, wasn't keeping pace. That was the premise for the original Inzata product, Inzata Enterprise, which is an AI-powered data warehouse and data management platform. It automatically models data together, puts it into a usable state, and lets users very easily create dashboards and reporting from it. It turns data into something that delivers value instead of sitting in a silo somewhere consuming cloud costs.
06:18Chris RafterSIGNLD is the next iteration of that thinking. Once we started to see generative AI technologies come out, it was the missing piece for us. But it turned out to be so different that we decided not to just make it a feature set on top of Inzata. We see a lot of companies out there throwing chatbots into their product, putting in a little chat assistant and calling it AI, but it's not really. One of the first things we looked at was security, because that's a big topic for everybody. And one of the first things we noticed about bolting on an AI chatbot is that it all relied on AI APIs. You'd go get an API from OpenAI or Anthropic, and they work really well. The issue is that our customers' data cannot go out over APIs like that. It has to be held securely.
07:04Chris RafterSo we architected a different approach. We run all of our AI in a captive walled garden inside AWS Bedrock. We're still able to run the most advanced frontier models there, like Sonnet and Opus and Fable, but the data never goes out over an API wire to another provider's data center. That lets us give our customers the assurance that they can use all the benefits and features of AI without worrying about where their data is going to end up, whether it's being sent to somebody they don't know or, worse, being used to train models, which we never do. That's a decision we made back in 2024, and we're very happy about it because it's come up numerous times, and it's been validated many times by customers as well as other companies in this space.
07:48Al Martin"Trust but verify" is probably the saying right now that couldn't be more important, and I would weigh the verify piece even more heavily than the trust piece. Anyway, on sharing data, I think that was a key decision, and I think you made the right call early. I have so many questions about what you just said. I noticed on LinkedIn, for example, you're the Chief AI Officer of SIGNLD, but you're still the Chief Operating Officer of Inzata. You referenced it a minute ago when you said a lot of people just add features onto whatever. Is this a whole different offering? Do you treat it as a different company? How does it work?
08:22Chris RafterWe almost treat it like a different company. There are some shared resources between them, but it is so radically different that we didn't want to constrain ourselves in the design process. We wanted to free our architects and designers to build what the market wants right now, in 2026 and 2027, not just a derivative of something we had before. So we're still improving both products, but they've forked onto two different paths.
08:50Al MartinI'm going to say this wrong intentionally so that you can correct me. Inzata is there for your data, to make sure your data is in one place, it's trusted, and so on. Then SIGNLD sits on top and does the data intelligence on top of that platform, using natural language via an LLM, so you can ask questions and do whatever you need to do on the Knowledge Graph of that data.
09:10Chris RafterThat's one way of doing it, but you can also use SIGNLD completely standalone. You don't need Inzata to use SIGNLD. You can bring data that was inside Inzata into SIGNLD, but what we're finding is people bringing data from all kinds of applications into SIGNLD: QuickBooks, NetSuite, Slack, their operational systems, HubSpot, Microsoft Teams. And the great thing about SIGNLD is that by pushing that data into a series of tables that are accessible to the Knowledge Graph, we can eliminate months and months of traditional data modeling, the setting up of data models and data structures we used to have to do when we built an analytics application for clients.
09:51Al MartinWell, how long does it take, then? I presume it depends on the size of the data, the clarity of that data, the cleanliness of that data. But if I said, hey, I'm going to go work with SIGNLD today and I'm going to work with Chris, what can I expect?
10:04Chris RafterSIGNLD is designed to work with a progressive delivery of data. What does that mean? People think they have to have all their data ready and cleaned on day one, and then a starter's gun fires and it's go time. SIGNLD is designed to meet you where you are and work with whatever data you have. What we've seen, with Inzata as well, is that it helps to give people a progressive approach. We have over 800 connectors, plus MCP connectors, to various data sources. Most people come in with one or two key systems they want to connect. Usually it's QuickBooks or another accounting system, or maybe an operational system or CRM data. They start with that, they start working with it in SIGNLD, and the Knowledge Graph starts to get built as they're uploading the data.
10:44Chris RafterThat's one of the other cool things: the Knowledge Graph is not something you have to sit there and build. It builds itself from its interactions with you. While you're chatting with SIGNLD and asking questions of the data, it's also listening to what you're saying, reading it into the Knowledge Graph, and then validating those parts of the Knowledge Graph back to the user. So it'll say, do you have seven departments or eight departments? Because when I read QuickBooks, it said you had eight, but you just told me you have seven, so which is it? It'll ask provocative questions like that, and then it'll remember the answer. It'll also remember which user gave it which answer, because believe it or not, people don't always give consistent answers to questions like that. The Knowledge Graph tracks that all the time, almost like a blockchain. We track every single change immutably, so you have a complete record of everything it was ever told, going back to day one.
11:29Chris RafterBut your question was how fast you can get up and running. You can be up and running within an hour and start getting questions answered about your data. It lets you come at your own pace and add more data sources, or connect some of your unstructured data sources to it. Everyone's path is a little different depending on their priorities.
11:47Al MartinSo it's not just a data lake per se. Again, don't let me put words in your mouth, because I don't want to mislead the listeners, but I heard lineage in there, so you've got lineage. I thought I heard governance in there as well. I didn't hear security, but you can tell us about security. You already said it's protected in the cloud. Can it be on-prem as well? I'll just shut up now.
12:13Chris RafterRight now it's available in the cloud, but there are parts of it that can be deployed on-prem, which we'll get to. One of the things we've also thought about a lot is what the future of dashboards looks like. We know how people have built dashboards up until now and how they like to consume them, and a lot of people are asking whether dashboards are dead. We think they have a great future ahead of them, but they're going to look different than before because of some of the benefits AI provides. So we're working with the concept of a living dashboard, or decision apps. Dashboards, at the end of the day, are code. They're code created by humans, usually with a tool like Power BI or Tableau, to lay out graphs on the screen and set filters and metrics. Underlying all of that is code, and we know AI is very good at generating code. That's exactly what we set out to do.
13:01Chris RafterSo we've been working on living dashboards inside SIGNLD. We've got a couple of them in beta right now. Going full lifecycle: you start using SIGNLD, you bring in a couple of your data sources, you start building a Knowledge Graph from them, and then you can start asking questions based on the data and the Knowledge Graph. But we said, what if we could use the Knowledge Graph, which is this wonderful repository not just of facts and data but also of the interrelationships, how departments work with one another, how business processes connect to one another and to the underlying data, to produce living dashboards that are almost like a customized web application for the employees and executives of the company? It lets them ask any question they want, it gives them daily updates, and it even lets them spawn work tasks and assign them to employees or to AI agents. We're really excited about bringing this out and what people are going to be able to do with it.
14:05Al MartinSo is this on the same level as a Tableau or a Cognos? Tell me how it compares.
14:13Chris RafterThe first thing is that Tableau and Cognos still require a lot of labor and time to set up. You have to design the dashboards, set up the data mart, the data lake, and all that. SIGNLD lets you sidestep a lot of that. You bring in the data directly from the source system, and SIGNLD does all of the data modeling and matching to identify the entities and their relationships within your organization. In the old days we called that an ERD, an entity relationship diagram. It would say Bob has a baseball and he gives the baseball to Sally, those kinds of pictures. SIGNLD does all of that in the background without you taking out a pencil and manually drawing those lines. It determines those relationships by looking at the data, the histories, and also the unstructured data inside the organization.
14:57Chris RafterI'll give you an example. We do a lot of RFPs at Inzata, given our public sector leanings. I asked SIGNLD to go back through our Slack and give me a recap of our entire RFP procedure, from when we first started doing them six years ago, how it had evolved, how many people were involved, what their roles were, and what they did. I was very impressed. Just by going through years of conversations and thousands of individual threads, it distilled a fairly accurate picture of our process for responding to RFPs over six years. It identified three generations of that process: early maturity, mid-level maturity, and a mature process, and it gave me a lot of detail around each. That's just one example of how much valuable information is contained in unstructured data sources.
15:53Al MartinSo you're not using RAG at all? It's all Knowledge Graph?
15:56Chris RafterWell, RAG had its time. We do use RAG, for example for parsing and processing documents on a one-time basis. But I think RAG has some deficiencies when it comes to what we're doing, which is giving people consistent, accurate, and traceable answers. One of the issues with RAG is that it's like a librarian you send into the stacks to get an answer for you every single time, and depending on the path they take through the stacks, they might bring back a different answer each time. That's something our customers weren't big fans of. They want consistency and traceability, especially with financial data. What was revenue last quarter? What's my profit margin on this product line? 98% accurate is not good enough. It needs to be 100% all the time.
16:41Al MartinI know SIGNLD is distinct from Inzata, but does SIGNLD come with Inzata if you buy it? Or if you wanted basically a data lake experience, would you have to get Inzata in combination with SIGNLD?
16:57Chris RafterThey're available as a package as well, but like I said, you can use either one standalone. They're not dependent on one another. We have people using just SIGNLD right now, and that works great for them. If there are parts of Inzata they want to use, they can do that as well.
17:11Al MartinWhat's SIGNLD built on? What's the technology underneath, beyond what you've already explained, like the Knowledge Graph? I'm just curious. You know what I'm asking.
17:23Chris RafterIt's custom, proprietary code for most of it. We use DuckDB to store some of the data as it comes in. We have a number of connectors we built ourselves, and we also rely on connectors from companies like Fivetran, which publish connectors we can use. So we have over 800 data connectors to enterprise data sources, and many of those come from the Fivetran relationship. The Knowledge Graph is really the center of all of it. We also have our LLM and chat features, and those are powered by LLMs running inside our walled garden at AWS Bedrock.
18:02Al MartinCan you use any LLM you want, or is that proprietary too?
18:05Chris RafterWe do have a selection of LLMs people can use, but only the ones offered by AWS Bedrock, because we have a no-APIs policy.
18:18Al MartinWhy do you have a no-APIs policy, can I ask?
18:21Chris RafterSure. It goes back to security. We're a HIPAA compliant entity. A lot of our customers store very sensitive data with us, and it's much more reliable, security-wise, to be able to say the data was loaded into our AWS cloud environment, and when a call goes to an AI, it's going to an AI running on a server literally right next to it, which returns the response. It's like running your LLM locally: you have your MacBook on your desk and a Mac Mini running an LLM right next to it, and the entire closed loop is your own computers. Your question goes to the Mac Mini, the LLM running on it gives the answer back, and you don't even need the internet.
18:57Al MartinI like how you position safety, and the fact that you're not sharing data. The data resides within the client's environment, presumably single-tenant. But let me ask you this. How do you ensure end-to-end governance? Your guardrails are up, you're protecting the data, encryption, all that, but you're also protecting against prompt injection, that kind of stuff. What are you using to make sure end-to-end security is paramount?
19:21Chris RafterWe have multi-factor authentication and Auth0 on the login side, so you have to log in to your SIGNLD environment. Everything in the environment is encrypted, especially the customer data assets. In addition, the Knowledge Graph, which is expressed in a series of documents that describe the rules, policies, and metrics that make up the customer's Knowledge Graph, lives in that same environment and is also encrypted. We try to limit the number of actors in the sequence. When you're working in SIGNLD, you log in, and your data is housed in the same place where the LLM answering your question and reasoning toward your answer is running. It's the exact same environment, not going out over an API to a server someplace else that we may not know.
20:10Al MartinDid you help build this personally?
20:11Chris RafterOh, yes. I prototyped it and did some of the coding, but most of the credit goes to my dev team for making it a reality.
20:20Al MartinI can tell, because you're very knowledgeable. You can tell the people who are really in the code from those who are adjacent to it. Clearly you know what you're talking about. Before I move on, I wanted to ask one more thing related to a previous question. I asked what problem you saw in the marketplace that brought you here. The follow-up is: was there a particular customer experience or light bulb moment that convinced you, when you already had Inzata, that you needed SIGNLD? That you had to take this next step? What was the light bulb?
20:46Chris RafterMost definitely, there were a couple. One I remember very vividly was in early 2024. Inzata had reached the finalist stage with an organization looking for a data warehouse and dashboard platform. We were a good fit for them, and they did end up going with us. But I remember in one of the final presentation meetings, we were demoing to one of the people in the room how you build dashboards in Inzata, and she said, "Well, why do I have to do that?" I said, "What do you mean?" She said, "Why do I have to move the mouse around on the screen and drag and drop the widgets into different places? Why can't I just use AI and tell it what I want?" And I didn't have a good answer for her that day. The answer at the time was, well, this is the way it's built. But when I went back and thought about it, she was right. Why should she have to do that?
21:34Chris RafterThat became a major motivator, and it also told me something about the market. Market expectations have shifted dramatically in the last 18 months, and we want to get ahead of that and be where the customer wants us to be. That's why we started prioritizing the development of SIGNLD back then. We saw a time coming when people would see the automation and force multiplier effect from these technologies, and they'd come to expect them from their everyday software tools. There's almost always something that serves as a trigger.
22:06Al MartinThat's a pretty good story. I want to key on the data here for a second. As you mentioned, everybody's talking about AI agents and large language models. But what I find interesting, and I have my own point of view on this but I want to get yours, and the good news is I'm the host and you're the guest, is that most companies still struggle with getting their underlying data in order. In your experience, are we building the future of AI on a shaky foundation? A shaky foundation meaning the data is a mess. Is that your thesis?
22:37Chris RafterYou're about two years too late with that question. If you'd caught me 24 months ago, I would have said absolutely, the data has to be pristine and clean and people really need to worry about this. I would have really embarrassed myself. My mind has since been changed 180 degrees, by evidence. What I've seen is that AI is incredibly agile and adept at navigating less-than-perfect data. One thing we found with SIGNLD is that we can take in unstructured data, dirty data that's been that way for the longest time, and the AI is still able to make sense of it despite all of those sins, if you will. We've been very pleased with the results. I had my mind completely blown and changed over the past 24 months seeing what these AI technologies can do.
23:26Chris RafterAnd we took that a step further. In SIGNLD, once you get past the Knowledge Graph part and start asking questions about your business, it's not making an AI call out to an LLM and using RAG or something like that to come up with an answer. The bulk of the answers it gives you, especially in the living dashboards, are deterministic. The data is fixed and the metrics are fixed, and that's what gives the experience its consistency. If you ask a question on a Tuesday and ask the exact same question on a Thursday, you get the exact same answer, as long as the data hasn't changed. We've been very pleased with how the AI has turned dirty or less-than-perfect data into reliable answers.
24:07Al MartinThe funny thing is, you took my follow-up question with your answer, because I was going to say, wait a second, isn't AI advanced enough to take less-than-perfect data and take a probabilistic approach, so you don't have to clean up the data like you used to? I think you answered that, and I do agree with you. Now, it depends on a lot of factors. You have to have enough data to get the same answer, as you mentioned. If you're looking at a very small dataset and it's incorrect, you might have trouble. But where is the line for you? It sounds like you're very experienced in data. I'm a data guy as well. When do you say, hey, we're going to have to clean this data? What tools do you use to look at it? How do you guide a customer? The last thing you want to say is you're going to have to spend a ton of money to clean all your data before we can do something with it. That usually doesn't sit well. I think customers expect it, but they won't engage in it from a cost perspective because it's going to cost them a fortune. What's your guidance?
25:12Chris RafterI 100% agree that with traditional solutions, it would prove disastrous to try to build a data warehouse or a data lake on a shaky foundation like that. But what we've found is that the AI can reason past those kinds of things and also remember the anomalies, and those get embedded into the Knowledge Graph too. So the Knowledge Graph isn't just how we calculate revenue and profit. It also keeps all those little secret sins, like the records that begin with the letter Z that aren't real customers, they were test customers. Everybody has those. A normal ETL would just read those in and they'd be computed into your answers. The AI can say, if any record starts with a lowercase z and an underscore, that's not a real customer.
25:56Chris RafterI'll give you another example, with a story. We had a customer in the safety equipment installation business. They have technicians who go out and install safety products for businesses, like smoke alarms. For a certain brand of product, they had some employees certified to install it. I believe they had seven. Every year they'd go through and decide which employees were ready for an apprenticeship to get certified. Their operational system, the system of record for those employees, said there were seven. This customer was using SIGNLD, and the supervisor of the team was talking with it. He said, "Tell me the names of the people who can go do this certification." The system said, "You have eight." He said, "No, I don't. I have seven." And SIGNLD said, "I'm sorry, but I'm reading that you have eight." He asked why. Here's why: SIGNLD was reading the notes fields in their transactional database of all their jobs, and it determined that there was one younger employee whom the more senior technicians had been apprenticing and teaching to get ready for certification. It wasn't officially in their data anywhere, but the system read the notes fields and put together this fact from them. Needless to say, the user called me up. He was blown away. He said, "We never would have caught this. I was not aware my guys were training this guy." He checked, and they were. So he went and updated the system of record because of what SIGNLD told him from the notes fields.
27:30Al MartinIt just hit me that the name of the company, SIGNLD, is purposeful, it sounds like. I just talked to another client today about the signals in their data. You took the name. Say more.
27:42Chris RafterWe got lucky, yes. We lost a couple of vowels in there. It's S-I-G-N-L-D.
27:56Al MartinYou've got to do what you've got to do.
27:57Chris RafterThe domain name business, yes. But that did fit really well, and we were really happy with how it came together. We're riffing on that inside the product by giving customers the ability to set up Topics and signals. The Topics are the things they care about most, and signals are the things the AI can triangulate around those Topics. For example, if you have a keen interest in expanding to another city, or improving the profit margin of a certain product line, you can store that in SIGNLD and say, I'm Chris, this is important to me, I want you to remember it and think about it 24/7, and when I log back in I want to hear what else you've learned. So that's another feature, where it sends signals about the things you care about most. That's really what makes it good at decision support.
28:42Al MartinDoes this let the client ask any question it wants about its data, or do you have different modules built in, like an assessment of what's going on in the industry, paralleled with their own data? Do you provide anything else? Or is it, Mr. Customer, you're on your own, ask any questions you want, it's your data, I'll step back? Or do you add different prompts to help them make decisions on their data and give them suggestions?
29:11Chris RafterYes, absolutely. That describes a dream I've had for a really long time, because I learned long ago that all the data inside your company is definitely not all the data you might need or benefit from. So we wanted to make it very fluid for SIGNLD to work with the data inside your company, in a secure manner, but also give it the ability to go out, spawn agents to perform research on topics you care about, and bring that data back in. In Inzata we had a concept called data enrichments. Those were third-party data sets that could be brought into a customer's data warehouse and applied to the data they had: procedure codes for healthcare, pharmaceutical codes, US census and demographic data from the federal government. Those were all examples of enrichments. Unfortunately, they were all fixed. The only way we could have an enrichment was if we went out, got it, and put it in our platform.
30:03Chris RafterWith SIGNLD, we've opened that up dramatically. We now have agents in SIGNLD that can do research independently and search the internet, as well as other data sources and data sets out there. They could be publicly available data sets, or data gathered from other websites. That's opened up a whole new world of data sets we can bring in and make part of the customer's SIGNLD environment, well beyond what they'd normally think of as data they would ETL into their analytics environment.
30:33Al MartinAre these pre-canned modules, or something in SIGNLD itself? How does that work?
30:42Chris RafterThe enrichments we had in Inzata for a while were preloaded data sets that we would go out and get, curate, clean up, index, and make available. In SIGNLD we still have all of that, but SIGNLD also has the ability to proactively go out and seek data on behalf of the customer that's pertinent to whatever they're studying in their data set. For example, a number of nonprofits are using SIGNLD right now, and they have agents in SIGNLD that look at publicly available information and demographic data on different neighborhoods, zip codes, and localities, and bring it back into SIGNLD so the customer doesn't have to gather it on their own. It's there waiting for them, and it gets mapped to whatever data they have inside SIGNLD.
31:27Al MartinVery, very good. I've got several other questions I want to get your expertise on, some a little more general, not necessarily related to Inzata or SIGNLD, just your expertise in the industry. But first, what didn't I ask you that you wish I had, on SIGNLD or data, that you'd like the listeners to hear?
31:46Chris RafterI have to say, you were pretty thorough. I can't think of anything you left out. We talked about how it works. We talked about RAG and where the Knowledge Graph fits in. We see vector RAG as the first foundational layer, then graph RAG on top of that, and then the Knowledge Graph as a derivative of that, each one an improvement over the previous layer. The Knowledge Graph concept may evolve again in the future, and it is evolving very quickly. But right now we're finding that the Knowledge Graph combined with the deterministic layer is yielding the results customers are looking for. It gives them the easiest path to bring their data into SIGNLD using AI and build the Knowledge Graph, and then the Knowledge Graph gives them a deterministic experience where the dashboard always shows the same value consistently.
32:36Chris RafterIt also saves tokens, because a lot of people using, say, Claude or OpenAI are uploading their spreadsheets and asking questions, and that consumes tokens left and right. What we've done is use the AI to interview the user, evaluate the data, structure the data, and build the Knowledge Graph, but the actual queries are done deterministically and don't consume many tokens at all. Or we can use a very lightweight, low-cost model to do that, which is something else we're looking at. I do think there's going to be a token reckoning coming, with how much people are spending. Just like with anything, people are going to look for ways to reduce token spend and token usage.
33:17Al MartinThat brings up a question I didn't ask that I just remembered. How is it monetized? In other words, what's the purchasing plan a client could expect? Let me back up and ask it this way: how is it monetized, and what does an initial engagement look like?
33:40Chris RafterYou can start with SIGNLD for as little as $40 a month, and we also have a Free Forever plan, so it's very economical. Pricing is based on the number of data sources you have, the size of your Knowledge Graph, and how many operations you perform. If you have hundreds of data sources, a very large Knowledge Graph, and hundreds of users asking questions every day that hit the AI, that would be on the higher side. If you have a fairly basic five or six data sources refreshed daily and five or six users, that would be toward the lower end. It's based on a token budget each account is given. You get a certain amount for the month, you can be billed for overage, and you can set limits on that as well.
34:28Chris RafterThe other thing we've been working on, which I haven't seen anywhere else, is a value-per-token score. I don't know about you, but I use other AI tools all the time, and I get my bill with no idea which of my chats cost the most or which ones I got ROI on. So we built an ROI tool into the platform for the CFO, to see not only how much they spent on SIGNLD but what their employees got out of it. We built a sentiment analysis monitor into the chat, so the chats themselves can be monitored for value delivered. If an employee is getting answers to their questions, sharing those insights, telling the system it did a good job, that's all captured, stored, and put against the token budget they're spending. That was a request from one of our early executive beta testers. They said, I care how much I spend with you, but more importantly, I want to know how much value I got.
35:27Al MartinThat's ROI. Makes sense. So it's like a real-time ROI dashboard they get. I feel the same way. With all the customers I work with, you can't bill off activity or information gathering, because that's almost ubiquitous right now. It's all outcomes and ROI if you want to gain traction. Where can people reach you, and reach Inzata and SIGNLD?
35:50Chris RafterSure. You can reach me on LinkedIn. Just Google my name, but there's a Chris Rafter who's a popular country music singer.
36:00Al MartinYou don't do that too?
36:01Chris RafterThat's not me, no. If you get the country music singer, that's the wrong guy. Try Christopher Rafter on LinkedIn. Inzata is at I-N-Z-A-T-A dot com, and SIGNLD is S-I-G-N-L-D dot A-I. We've got trials there, and you can sign up. If you do a trial and you're not comfortable loading your data into it, that's fine. We have a data generator built into SIGNLD. When you sign up, you can tell it, hey, I run a marketing agency, or I run a veterinary office, and it will generate a small playpen environment with data similar to what you'd encounter in your business. That way you can try it out in context without uploading anything or connecting your data sources on day one. It's fully functioning, and you get to see all the features.
36:47Al MartinSIGNLD has been out for, what, a couple of months?
36:49Chris RafterYeah, a few months.
36:51Al MartinI'm sure you're going to tell me it's going great, but how has it been going? Any trials and tribulations you can talk about that others can learn from?
36:57Chris RafterOh, for sure. We've had people using it very intensely, staying up until 11 o'clock at night exploring their business and getting answers in ways they never had before. That was really exciting to hear. We also had some customers running out of context windows, so we had some of those growing pains, and thankfully we've taken care of that. Right now we've got clients coming in through various channels. We've seen a lot of adoption in the finance and professional space. CPAs and accounting firms have been using it quite a bit. But we've also seen usage from nonprofits, retailers, manufacturers, and real estate developers. So we're still figuring out which industries are going to get the most out of it, and we're watching and tracking each of them very closely.
37:43Al MartinWhat's your sense of the industry that's going to get the most out of it?
37:47Chris RafterSo far, CPAs have been benefiting a lot. We're hearing CPAs talk about saving upwards of 10 to 15 manual hours per employee per month, and eliminating software packages that cost them tens of thousands of dollars a year by moving those workloads into SIGNLD. We have a new flavor of the product coming out called SIGNLD for CPAs, which is basically an agency model. Say you run a CPA firm and do taxes and books for hundreds of business customers in your area. SIGNLD for CPAs would let you set up an account with SIGNLD and make the same SIGNLD experience available to your customers. They could upload their financials and drop in their receipts, PDFs, credit card statements, and bank statements, and it flows to you. You could then use SIGNLD to do their monthly books and give them insights and strategic advice about their business. They also get SIGNLD access themselves, to go in and explore their business and ask questions. So we're really excited about that.
38:45Al MartinThese are things I'm interested in. I'll finish with a quick rapid fire, but I'll ask one or two more first. There's an enormous amount of money being poured into AI. The way I look at it, NVIDIA is selling picks and shovels, Microsoft and Google are building platforms, and OpenAI and Anthropic are racing to build smarter models. Who do you think is going to make the most money out of this revolution? Where is the money going to land? I know that's completely different from what we've been talking about, but I ask other guests the same question. Go ahead.
39:17Chris RafterI used this analogy with some of my employees the other day. Do you remember the early dot-com era, when there were hundreds of search engines? Lycos and so on.
39:28Al MartinYeah, go ahead.
39:30Chris RafterHundreds. How many are there now? Like three? Google, Bing, maybe Yahoo.
39:35Al MartinWell, there's really one.
39:39Chris RafterSo I'm going to use that lesson to inform my answer. I think it's going to be the company or companies that figure out how to deliver the most value to users. You remember the story of Google: they stumbled across a special way of doing search, based on how many sites linked to another site, and that was unique. I think it's going to be the companies that figure out how to deliver the most value for the lowest outlay. I don't know who that is.
40:05Al MartinI'm not asking you to be a fortune teller. Of course, you can give me a stock pick if you want. No, I'm with you 100% on that. You could use so many examples. You could use fiber optics. When fiber optics came along, there were a few companies investing like crazy. They overinvested, kind of like what we're doing with data centers, and those companies don't exist anymore. It's whoever comes along afterward that picks it up at a fire sale. At IBM, that's one reason I say we're building the platform. We're not investing in data centers. I think the data center craze is out of control, and NVIDIA is pushing a lot of that. So it's going to be interesting to see how this all plays out. That's why I wanted your view. Here's a provocative one, and then I'll go to the rapid fire, related to your business, and hopefully you'll slam this one. If I open ChatGPT, upload a spreadsheet, and ask it to explain my business, why do I need an analytics platform? Like you said earlier, AI eliminates the need for perfectly clean data. Can AI eventually eliminate the need for dashboard BI tools and data analysts?
41:13Chris RafterI don't know that AI alone will do that. I do think Decision Intelligence tools that use AI will put a big dent in that, or largely eliminate it. What we're betting on is this: coding and building websites and apps has gotten cheaper than ever before in history. It's like when we used to print out a letter, put it in an envelope, and stick a stamp on it. Then email came out, that went away, and sending a message became basically free. The same thing is happening with creating a website or a dashboard. It's all code that AI can produce in milliseconds. So the price is going down, which means the quantity is going to go way up and the quality is going to drop. I think we have to go through that zone first.
41:57Chris RafterBack to your original question about dropping a spreadsheet into ChatGPT, a couple of things to note. Number one, when you open a new ChatGPT window, it has hardly any context, unless you're one of those people who edits your MD files and keeps all your personal context there, which you probably are, but most people aren't. So when you drop that file in there, it doesn't know anything about your business. It only knows what's in that file and what you told it. There's a lot of talk about context being the most important thing here, and we agree. What our Knowledge Graph does is let you very quickly and easily extract the context your business needs to get the most out of AI, put it in a place where it's safe and secure, can be maintained, and is auditable, and then use it with any AI model you want. ChatGPT plus a spreadsheet gets you to one destination. ChatGPT plus a SIGNLD Knowledge Graph connected via MCP, where ChatGPT can check its work against the Knowledge Graph you developed in SIGNLD, is going to get people a lot further.
42:58Al MartinMakes sense. I appreciate it. Nice answers. You were ready for it, but he didn't see these questions, folks, I promise. All right, here's the rapid fire. There are no right or wrong answers, because sometimes I'll ask who's going to be the AI winner or something like that, and we know that can change. We're not asking you to be a fortune teller, but I do want your view. You're an expert in the industry, so let's use it. Quick answers. The AI winner: OpenAI, Anthropic, IBM, Google, Microsoft. Who has the biggest long-term advantage in your mind?
43:33Chris RafterIt changes weekly. I don't know. I do think there's going to be a backlash on token consumption, but I don't know what these companies' plans are or what their future economics will look like. So I don't know.
43:47Al MartinThe biggest threat: is AI most likely to disrupt Microsoft, Salesforce, Google, or traditional software startups? Who's it going to disrupt most?
43:57Chris RafterIt's definitely disrupting software companies. I think the companies at the very top are going to be relatively safe, and the ones at the bottom developing niche apps, like a CRM for lawn mowing businesses, are going to be relatively safe. The ones in the middle are going to be impacted most: companies in the 10 million to 10 billion range.
44:21Al MartinHere's the slam dunk, the data problem. What's worse for most companies: bad data, too much data, or data that nobody can access?
44:27Chris RafterProbably too much data, more than they're able to process. But I'll say the biggest risk is not thinking broadly enough about what AI can do for them. That's probably the biggest mistake I see people making now. They say, I'm going to implement this, they get to a point, and then they stop. I always ask, why'd you stop there? Why can't you push it further? They say they didn't think so, and then we say, well, here's what you can do, and they say, yeah, you're right. So we're constantly encouraging people to expand their thinking and expand their expectations, just like that woman did for me two years ago when she said she didn't want to make dashboards with a mouse. I have to thank her. She should get some residuals.
45:11Al MartinAll right, let's talk about skills. If you were starting your career today, what would you learn? Data science, software engineering, business strategy, AI?
45:17Chris RafterI think strategy, and customer-facing skills are extremely important. I love the concept of the forward-deployed engineer. We've been developing that role at our company for quite a few years. We also just became Anthropic certified. We got all our full-time employees Claude certified. That kind of skill, being able to work with someone, have enough business strategy training to quickly understand their business, and then go in and help them with their problems, is never going away. It's always going to be needed. That's where I would start today.
45:50Al MartinWhat technology or industry buzzword do you think is getting way too much attention right now?
45:55Chris RafterThat's a good one. I'm going to have to think about that. There are so many out there. I can remember a time when data centers were at least innocuous. There's been a lot of backlash against data centers, ironically on social media, which wouldn't exist without data centers. I think there's still a lot of understanding to be spread around about how these actually work. The risk of every small town getting a data center in its backyard is probably not realistic.
46:31Al MartinI worked in data centers for years, and most of them are where the fiber is. So I say to people, is there fiber on your street? No? Well, guess what, there probably can't be a data center there. Underrated technology: what's not getting enough attention?
46:51Chris RafterI would say GitHub, and the amazing development going on there. The number of developers on GitHub has exploded, and there are some great tools on there. Two or three years ago, we used it, but I didn't spend much time there. Today it's like going on Facebook used to be. There's something new there every day. I just saw that somebody vibe coded a way to update the firmware on your smart TV so you don't have to look at ads every time you turn it on.
47:24Al MartinI like that. Who wouldn't? I'm going to have to investigate. This is why I do these questions. What's the first white-collar job or business function you think AI is going to fundamentally change?
47:36Chris RafterI would say people who aren't communicating and are just performing the same process over and over without looking for continuous improvement or trying to deliver more value. One of my favorite sayings is that it's not that AI is going to replace your job. The bigger threat is someone who knows how to use AI well replacing you. We're seeing AI become an extremely valuable partner for nearly every role in our company, and we're setting people up with the knowledge, skills, agents, and token plans to make the most of it. I've been extremely impressed with my own team and how they've developed. We've automated work without any job cuts, but with a major elimination of manual tasks, moving things like proposal production and product management into AI tools. We've also been able to eliminate a couple of systems we were spending money on that we don't need anymore. All of that is producing positive results without impacting a single job.
48:31Al MartinThat's awesome. Only a couple more, then we'll call it. You're doing great, by the way. If I gave you a million dollars and forced you to invest it in one technology over the next five years, what would you invest in?
48:41Chris RafterI really like SpaceX's concept of putting data centers in space, but I think we may be a ways off. I'm not going to say anything about Tesla, but Starlink is a great technology, and it does have the promise to change the world positively. If you look at the big vision for SpaceX around putting data centers in space, that's probably the one thing I've heard that could provide major benefits for humankind. It's also probably the least likely.
49:16Al MartinTen years from now, and I know that feels like forever today, so it's not even fair to ask, what's going to be commonplace in business that most people would find hard to believe right now?
49:28Chris RafterI think we're going to continue to see the removal of friction, especially friction we're used to and just put up with right now. There are things we did 20 years ago that we don't do anymore, and they quietly fell by the wayside. I think we'll see a lot of that kind of friction removed. You've got coding going to almost zero, and I think decision making is also going to become much cheaper. It won't take as much to get higher-quality decisions, and that's going to benefit many fields and industries. The continuous removal of friction in every process and every transaction is going to accelerate.
50:08Al MartinMakes sense. All right, man, I think we're just short of ten, but I'm going to stop there. You answered those great. It was like I prepped you, and I promise, listeners, no prep whatsoever. It's the first time I've met Chris. Chris, I'll give you the last word. Anything you want to say to the listeners before we sign off?
50:23Chris RafterI just want to say thanks, to you and your listeners. I really enjoyed this, and I appreciate the opportunity.
50:28Al MartinYou're a good man. I wish you the best. It sounds like you've got a great strategy around both Inzata and SIGNLD. Listeners, you know where to reach him: signld.ai. He's pushing it, as he should be. I wish you the best, my friend. You had great answers, and I can tell you're in the code, which is what you want a founder to be. So thank you for being here. Listeners, thank you for listening. Reach us at almartintalksdata@gmail.com. We'd love to hear from you. I'll see you on the podcast. Bye-bye.
About the guest
Chris Rafter
Chief AI Officer, SIGNLD
Chris has worked in data for 25 years, building data warehouses and analytics programs for private companies, state agencies, and K-12 districts.

What could Amendment 3 mean for Florida renters?
By Romy Ellenbogen
The Tampa Bay Times covered an interactive tool the SIGNLD team built from public data, which lets residents change the variables and see the result for their own city.
Read in Tampa Bay Times“There's many, many variables here and that's exactly why we built the tool.”
Media inquiries
Chris Rafter is available for interviews on Decision Intelligence, Knowledge Graphs, AI governance, and putting AI to work on business data.
ethan.bellows@signld.aiAbout SIGNLD: SIGNLD connects 800+ business systems into a live Knowledge Graph and answers questions from your own data, with every answer traced to its source. SIGNLD is built by Inzata Analytics in Tampa, Florida.