# From Raw Data to Trading Signal: How AI Is Changing Market Access | Future Finance Forum 2026 - Channel: [ETHSofia](https://streameth.org/ethsofia) - Date: 2026-10-06 - Duration: 31:07 - Topics: 4PTO Labs, AI agents, AI trading, Blockchain Week Bulgaria, Dragomir Dikov, F3 2026, Future Finance Forum, Jordan Duclos, Jungle Trade, Sally Meouche-Ghrawi, Truflation, macro data, market data, prediction markets, real-time inflation, regime shifts, trading signals, Science & Technology - Watch: https://streameth.org/watch/yt-1rchJBIJH3w - YouTube: https://www.youtube.com/watch?v=1rchJBIJH3w ## Description With Sally Meouche - Ghrawi (4PTO Labs, moderator), Jordan Duclos (Truflation) and Dragomir Dikov (Jungle Trade). The last session of the day follows the chain from raw market data to a trading decision. Jordan Duclos explains how Truflation turns around 100 million daily data points into some 500 live indexes, arguing that monthly government inflation figures are too slow and that in finance latency is death. Dragomir Dikov of Jungle Trade argues that market noise is worth analysing rather than discarding, that every signal should carry an explicit confidence level, and that full trading automation is inevitable but premature today. Moderated by Sally Meouche - Ghrawi, the panel debates where the edge lies and whether shared AI models lead to correlated strategies, with Dikov naming trust as the scarce resource in three years and Duclos betting on real-time data as markets move to 24/7 trading. Panel at Future Finance Forum 2026, 25 September 2026, Sofia. Part of Blockchain Week Bulgaria 2026. Speakers ▸ Sally Meouche - Ghrawi, Founder and CEO, 4PTO Labs (moderator) Sally Meouche-Ghrawi is Founder & CEO of 4 Point O (4PTO) Labs, building across AI, fintech and Web3, and Co-Founder of Buildrooms, Bulgaria’s first Web3 hub. She works with startups, protocols and enterprises on emerging tech, product development and market strategy. LinkedIn: https://linkedin.com/in/sallymg ▸ Jordan Duclos, Head of Growth, Truflation Jordan is Head of Growth at Truflation. After starting his career in SaaS, helping banks and public institutions improve customer engagement, he moved into blockchain, working across the Ethereum and Bitcoin ecosystems. At Truflation, he focuses on making inflation and economic data more accessible through clear, impactful storytelling. LinkedIn: https://www.linkedin.com/in/jordan-duclos-b48b72153 ▸ Dragomir Dikov, CEO, Jungle Trade Dragomir Dikov is Co-founder and CEO of JLabs, the team behind Jungle Trade, building trading technology and market-analysis tools. He holds a master's degree in Structural Engineering and has 15+ years of experience in geotechnical modelling and structural design. He is a Python programmer and quantitative data scientist focused on probability and stochastic modelling. LinkedIn: https://www.linkedin.com/in/dragomir-dikov Chapters 00:00 Introduction: from data to decision 00:57 What has been missing in market data? 02:43 Is market noise useful? 03:57 Real-time inflation data 07:16 Regime shifts 08:45 Signals and confidence 09:54 Where is the edge? 14:43 From analysis to autonomous trading 16:25 Prediction markets as a signal 20:16 What if everyone uses the same models? 21:41 Three years out: where will the scarcity be? 25:54 Q&A: Can human traders compete with machines? 28:35 Q&A: Buy data and signals or build your own? Blockchain Week Bulgaria: https://www.blockchainweek.bg ETHSofia: https://www.ethsofia.com Future Finance Forum: https://www.blockchainweek.bg/f3 Follow Blockchain Week Bulgaria X: https://x.com/BWBulgaria LinkedIn: https://www.linkedin.com/company/blockchain-week-bulgaria Follow ETHSofia X: https://x.com/EthSofiaBG LinkedIn: https://www.linkedin.com/company/ethsofia Telegram: https://t.me/+b-33LJUpAB5iODNk Nothing in this video is financial advice. About the organiser Blockchain Week Bulgaria, ETHSofia and the Future Finance Forum are organised by the Bithope Foundation, founded in 2014 by Vladislav Dramaliev. Inspired by Andreas Antonopoulos, it is Europe's first non-profit operating exclusively with bitcoin donations. Over more than ten years, it has supported 50+ charitable campaigns, and in January 2016 it co-founded the Sofia Crypto Meetup, now the region's longest-running monthly crypto event. https://bithope.org ## Transcript Hello everyone and thank you for putting up with us I would say on the on the last session of this day. Um when it comes to I would say AI and trading there's always no shortage of the discussion around it but what is more interesting at this point in time for us to tackle is what happens actually like what is the process and the chain before the actual trade and um when it comes to trade I mean there's always the the data and then there's the the the understanding of the data and then there's the context which leads ultimately to the to the signal and in the end to the decision. Uh so we so we have with us today uh Jordan from Truthflation and Drago from um Jungle Tate. Uh I would say the first question maybe to to Jordan to start is that when it comes to I mean data there's always I would say no shortage of data. But what what has been missing up till this point? Is it I would say um like the speed of of of processing this data? Is it uh the quality of the data? Um yeah, I will leave this answer to you basically. >> Yeah, I mean first of all, thank you to the organizers for having us some sponsors. I think the number one thing it's not that it has changed, it's economic intelligence. I think for whoever is in the audience, whether you're a trader or it's your first time in crypto, at the end of the day, where do you get your information? Is it on the news? It's time to buy gold. Is it on Twitter? Oh, maybe you should go look at silver. Is it uh you know through your network going to events or is it using AI tools? And so I think that's the one thing that's changing a bit more with AI is that now people are starting to ask more AI in terms of insights of what's the impact on commodities of XY Z. >> What about you Drago? Do you have anything to add on that maybe? >> Well, I think that uh we have a large quantity of data nowadays. Can you hear me? uh a large chron quantity of data these days we definitely need some kind of new generation tools for example like AI to be able to help us contextualize this information. This is not easy nowadays because uh you have uh how Jordan mentioned uh data bombarding you all over the place with uh you know different statist statistical information also and uh using specialized tool to be able to contextualize this I think that is is going to be very beneficial >> okay so then what happens with the noise >> yeah noise is it's very interesting subject as actually it's people are struggling with noise in data for centuries This is not new phenomenon. We have noise in for example early radio transmissions. And uh there is definitely noise in the market. But uh let's ask ourself one question. What is actually noise? What we mean by that? Well, this for my humble opinion is uh some u element of the market that we cannot explain. But is this mean that if it's not explainable, it is not useful for us. It is not useful to be able using these new generation tools to be able to explain this phenomenon and to be able to capture any dependency structures that it may have with the with in the market. Uh in the end of the day, it's actually part of the market. I think that uh the today day and age give us the the the the proper tools, the proper machines, the proper knowledge to be able to actually u analyze it than just to to try to eliminate it. And Jordan, I think when when we last spoke, you mentioned that basically throughflation was built around the the point of like latency when it comes to economic data. >> Yeah, it's I mean, I don't want to go into cliche sentences for the audience, especially at the end of the day when everyone's tired. Um, but that being said, I I do believe in actually the signal in the noise, which you know is a great expression. And we have 100 million data points that we analyze every day. But obviously nobody cares about 100 million data points. And so we've transformed these 100 million data points into actually 500 live indexes such as the US inflation economy. And so that's why like we have Goldman Sachs that's you know wants our data. We have Wellington. We have Bridgewwater. We have the Federal Reserve government. And these were things that a couple years ago everyone thought that we were crazy. And today everyone wants our data. So I don't know if we can say it's a signal in the noise but we are basically trying to make dry data. We're trying to make it into insights because hopefully you know if you're in the audience if you're a trader as I said or if you're a new person at the end of the day you want the alpha you want that information that you can use to make a better trade. And so where do you get that information? I'm not going to shield you know necessarily translation or jungle trade but those are avenues where you can get that signal. um to make better trades generally speaking. >> So who's consuming this data? You mentioned a few names a bit ago. Who's consuming this data? And does this basically bring us to the to the point of like contextualization here? Uh because we did mention that there's I would say hundreds of different data points but each and every like the audience is different and then the purpose of this data is also different to each. >> Yeah, I I mean I I can take it. Um, there's one of the most important indicators on the planet and that's inflation. And a lot of people don't know, but there's over $130 trillion attached to inflation. And so if I tell you 3.4, 3.5, nobody cares. Nobody understands. If I'm telling you that your rent is going up from $700 to $1,000, that's going to speak to you. And so you have a number that is US inflation that is given only one time per month by the US government which is crazy because we live in the day and age of social media of instant gratification. So why wouldn't you be able to have daily access to what the inflation is? And so that's something that we've been working on for the past few years is how do we provide real-time data on inflation? And as I said earlier, at first people thought we were stupid and crazy because we were competing with the government. But today we have the government that actually is working with us and wants our data and the Goldman Sachs and everyone actually wants that. So there is value in it because if you understand where the economy could go cuz I'm not going to pretend we know for sure then you can go on the stock markets because behind inflation there's commodities. Everyone cares about gold. Everyone cares about gasoline. Yep. >> So what are you buying? How to use that data and so >> Yep. >> Yeah. I mean this is a good point because he you you touched upon I would say like price movements and price shifts uh and predicting those movements and I think when it comes to Drago uh I know that you're interested in regime shifts >> regime regime shifts. >> Ah okay okay. >> So how does this basically let's say how does this translate into regime shifts? So yes, I I think that uh definitely inflation is contributing factor to the regime change and uh Jordan here making is making an excellent point. Uh providing good quality and reliable information in the decision making making process is something very important and uh they actually you know um are working on subject that is extremely important like inflation. So inflation and there are can be many contributing factors to the regime change and this is the actually the main focus on my work uh uh today and now I am actually working on uh several projects that are involved in detection of possibility of uh contextualizing the possibility of of of a regime change and one of the contributing factors one of the contributing data sources is uh definitely going to inflation. So the the the good quality data is something that we need to strive for and uh we need to be able to use in our decision making process especially if we are going to investigate something complicated as a regime change >> and do you think that we're moving towards a direction where I would say like a good kind of signal becomes like what what is the persistence of this signal? Yep. >> Can it become dynamic? >> Well, signal is very broad term for for conclusion of our observation. Uh signal definitely cannot be one thing alone. Uh it can be multiple things or aggregation of multiple things. And uh when we come to signal my my at least my approach is that any signal that is derived from raw data need to be um penalized in terms of confidence. How confident are we on this uh decision based on the evidence that we have uh gathered. So it's it's it's basic science approach. We have an evidence we have some conclusion. we have an evidence behind it and we have some certain amount of quantification like probability that this decision is accurate or not >> okay so where's the edge is it I would say is it the better data is it the faster data is it the better models is it the is it the execution >> um yeah I mean where's the edge >> I think there there's two parts depending in the audience on whether Once again, you're a fun or a trader. The the reality is you're never going to be able to beat uh the team of, for example, Black Rockck. They have their own models. They have some of the smartest analysts on the planet. They have data that they're paying in like a fraction of a second they have it. And as you can see, not to go political, but even with the social media of Trump Truth, some people were paying $100,000 for the API because they want to have one second earlier that he's going to mention, well, Dell is looking interesting and boom, the stock goes up, right? And so the reality is that you're probably, unless I could be wrong and you're in the audience, you have that alpha, you're not going to beat Black Rockck, you're not going to beat Goldman Sachs. >> However, using the right signals, uh, you can actually trade more the mid to long term of like understanding >> obviously and I'm a bit biased in our case, it's inflation, but inflation is tied to everything else, right? So if you have an idea where inflation is going to go the next month or in the next year then maybe you want to buy bonds or treasury maybe you want to buy certain commodities. So if you understand where oil is going or gasoline and what's the impact on housing market, what's the impact on food and so this way you can take positions on things that you really believe in and that could obviously bring you profit over time and I'll give one example that I think a lot of people are not talking that much about surprisingly is everyone is obsessed with AI for very good reasons and I'm also part of that train. Uh but not that many people are talking about robotics. It's just a general topic. You see it on the news, China just organized, you know, the the tournament uh for robotics, but think about how robotics could impact even the economy from a productivity standpoint. And so then you look at the stock market, whether it's China or the US. What are the companies that are going to be impacted? Nvidia just released its report like a month ago. Uh they did a 96.2 billion in revenue just for this quarter and the projected revenues are 600 billion next year. What are the companies if you look at the top in the world and you look at the trickle effect of if you think that there's a trend with robotics what are the stocks that are going to be impacted potentially how is it going to impact the price of food you know so that way if you're a trader you can have an idea and insight of like maybe I need to short the market here maybe I need to long the market here maybe I need to buy this and this going back to the topic this is why AI is so important is that once you start seeing a bit what's happening in the world. Nobody has a crystal ball. But you can then use AI to confirm this thesis. And we were discussing this earlier like like a scientist. You have a hypothesis and then you need to really test it out. Is it going to hold? And so that's where you need once again the data. >> Yep. Yep. Yep. I definitely think so too. Yeah. >> So, so what about you, Drago? What's your take? Where's Where's the edge? Where's the edge? I mean, what's your take on that? Well, well, the edge is more like u if you it's a complicated question where is the edge you know it's it's not actually a prediction it's more of understanding you know uh if we have for example as Jordan mentioned a boom of AI and a boom of robotics this is something that we need to think about in order how is this going to influence the market because at the end of the day market is not in vacuum. It's not isolated thing you know it's it's um relying on people for now at least uh we we cannot say for the future but for now it's much relying on people and actual people are participating in these markets. So in in in worst for example macroeconomics conditions where people are uh without jobs where robots are for example taking their place the market is going to be whole different and uh where is the edge? Uh I think that the edge is the the better understanding >> at least on my point of view. >> Yeah. >> Uh because I use AI on daily basis. Uh I I confess that uh in the beginning I was very skeptical, skeptical uh even my colleagues uh needed to how to say to to push me to that uh you know using AI more often uh in my process because it make my work more efficient uh in the end of the day. So I definitely think that they can help us understand something that we previously uh did not have the capacity to understand and and I definitely think that this is the edge. >> You you mentioned I would say like an important sentence in your line when you were answering which is like basically we're still relying on people. So this actually brings me to my my next question like do you think that or are we moving uh with AI from analysis into let's say autonomous market participation? [clears throat] >> This is like >> this is tricky question >> like [laughter] like what happens after analysis? Well, I the the the odds if I if I need to to for example take an take a bet on it. Uh I will definitely think that uh the automation using AI in different processes even in the decision making process is inevitable. It's not if it's going to happen, it's uh when it's going to happen. At least on my point of view for now we are too early for this stage. I think that we can uh definitely take it as a step decision in in in this implementation. Uh as we all know uh AI is not perfect. AI make mistakes. I experienced it uh in my own uh research. So we need to be careful if we are going to delegate uh the hold the automation for example in trading and uh definitely definitely I think that we going to and there is evidence in this direction that uh uh we need to slow down a little bit and uh you know make a little bit more research you know and uh then we continue with fully automation. >> What's your take? I agree but I disagree also. [laughter] >> Okay. >> It's um I guess going back to know the practical examples and you know obviously crypto and blockchain. One of the best places to find signal today is actually prediction markets. You know whether we like them or hate them. Um there's definitely a strong product market fit. They've raised at a billion dollar valuations and everyone's starting to use them more and more for sports and other things. Right. So on one side it's a bit controversial controversial to say this but it is a lot of gambling. However you can also have signals when the question is very important and interesting. So will the interest rates uh rise in the US that one once again that's tied to trillions of dollars. So that's an interesting answer or US elections to see what people are thinking, right? And so going back to the use of AI is where can you find all these signals and there's multiple places and in our case and I don't want to sound like I'm kind of shilling and and selling our product but we we've built our own proprietary harness or macro agent which is specialized on the US economy. And so this is a tool that literally you know we have the researchers from the Federal Reserve in the US that are using our tool. We have Goldman Sachs as I said all the ones before. We have Kathy Wood of Arch Invest who just wrote actually in the investor letter she talked about true inflation like two days ago. And this is a tool that we've actually made for retail because you can go there and you can ask simple questions such as what's the correlation between inflation and the price of gasoline and diesel. And if it does not have the answer, our agent will tell you I do not have evidence. So it's fully evidence and sourced back. And I think that that's something that is so important today is that whether you're going on social media or the news or talking to your neighbor, friends, family, everyone has an opinion, but very few people are able to back it with actual facts. And so in our case, our AI is not giving an opinion. It's just giving facts. It will tell you, okay, this is the price of gasoline on the month. This is what it was last month. This is how it correlates to inflation. Inflation was this today. It's going to be this tomorrow. And so it's going to give you the tools to once again once you use RAI for example then you can go to prediction markets and see what it's saying then you can go to Twitter or X maybe X is going to confirm or say the opposite and so final example is that when you look at our data researchers tend to agree with us interest rates shouldn't have gone up in the US but when you go to prediction markets 90% of people thought it would go up and they ended up going up but our data most of the real time data said it shouldn't have gone up. And so that's the difference is that sometimes what happens is what shouldn't happen. But you also have to be able with AI or as a trader to understand how the market is going to react because sometimes it reacts in a way that doesn't make sense. And I'll finish on this example is when the interest rates uh in the US went up uh last week um the stock market crashed when it happened which is what is supposed to happen. However, Bitcoin went five or six% up and that is something that is inexplicable. You talk to any expert, >> whoever tells you there they know the reason is lying. The the reality is that there's market makers in wells that are actually unknown. They have so much money that they saw an opportunity because everyone thought that the rates would go up, that the stock market would go down and that Bitcoin should go down. So, they saw an opportunity and they make a fortune. And that's the part where AI is not there yet. humans are not there yet because you're not working at Black Rockck or at a place like that. So, it's just how do you find that edge? You know, there's no perfect answer. There's no easy way to make money, but still by using these tools, you're able to get an edge to win on the midterm or short term if you're very smart about it. >> Yeah. Yeah. And the thing is also like once if everyone's using kind of the like similar models uh like but then isn't there some kind of like a correlation when it comes to the strategies? >> No, >> they can be correlation between strategy if you using the same model but uh using the same model doesn't necessarily guarantee the same outcome. >> Mhm. >> Jordan just made an example of this. you know, we're looking at the same model. The model predictions is this one, but the Bitcoin is going in the other direction. So there is a >> but but then comes the the macro I would say but but using you know we we are using the same tools for decades now. >> Yeah. >> And we definitely didn't produce the same outcome for everyone. >> It's uh it's more complicated. It's about uh how we combine these tools. How are we using them? What information we drive from these tools? What is our decision making process? Which correlation structures we observe more often? Uh which is the what is our risk strategy? What is our portfolio management? So it's complicated ecosystem not just uh you know uh using for example an AI agent to be able to contextualize some uh for example data streams. So I mean we're we're reaching closing and I would I would end it with kind of a more forward like looking question. So imagine if we're sitting here 3 years again agents are much more capable. Uh the data >> is more I would say market data is more like machine readable and the costs of of all of those I would say processes have fallen significantly. Then what what is like where where's the scarcity? I mean what is then the is it is it the data is it the the trust is it the execution is it the interpretation like 3 years from now what is the I would say where where would the edge be or where would the scarcity be three years from now >> if I have to pick one it's going to be trust >> trust >> most definitely yes uh we are going to produce more data given what we did uh what we do here with Jordan we are producing more data essentially. uh so I think that uh the trust is something that we need to take into account and these actually um gaining this trust actually comes not from only from data quality uh that is data is proven in terms of uh uh what is his usefulness uh what is the problem that is uh going to solve uh and which are the people are that going to to use it and um um I definitely think that trust is something that We need to establish in the very early stage and some of that trust is going to be uh gained by disclosure of our uh ways of our principles of ways our our models work. What are the advantages of our model? Which are the disadvantages where our model is valid? Where it is not valid. So essentially this is the information that is going to be uh you know uh essentially more important for the end user. Um yeah sorry I have a problem with the mic. Um, I just want to throw a couple, I guess, thoughts out there, which I don't want to make it sound like it's BS or or marketing in general, but my take and I think it's also what we share at True inflation is that in finance really latency is deaf. If you're late, you're already losing money, right? >> It's a fundamental truth. And so if you think this way, then the conviction is that literally in the next year or two, markets are going to be moving to 24/7. Meaning that real time will be critical. And this is actually something that's been confirmed with the Federal Reserve chairman in the US, Kevin Worsh. As soon as he got in, he said, "I'm doing a task force about inflation about real time." And so I I'm talking on a weekly basis with funds and traders and at the end of the day the only thing they care about and that's also why they want our data is they want real time because even though data there's no certainty but if they can get the real time possibility of where the market is going already for that for them that's worth a fortune. We're not smarter than Goldman Sachs and Black Rockck, but they're using our data because they see value in getting a slightly different take and also they don't have access to real time. And to finish on the overall vision, >> I think that something we need to take into account and I forgot the official numbers and statistics, >> but I think it's 70 or 90% in the next two years that agents are going to be trading. And we're we're still focused a lot on humans trading traders. But the reality is that as AI grows, agents are going to be trading potentially better than us humans. And so that's something >> and 24/7. >> 24/7. Yeah. and they'll probably make smarter decisions because they'll have the whole context and they'll understand better a topic. So, question mark. I don't have the answer, but that is, I would say, very likely to happen. The question is, how do you profit and make money off of it? And if I knew, then you can find me on an island in a couple months, right? >> Thank you guys for your time. Um, do we have questions? >> Because I see we have three more minutes. So, yeah. So these are some of the questions. With AI trading bots advancing so quickly, how long can human traders continue to compete with the machine? How can you what? >> How long the humans traders will be >> how can humans compete against the machine? basically what we what you touched upon the last point which is agents being there 24/7 having more context and more speed at this point in time >> I I would say still the quick answer is you need a trade with AI but you need you can't be dependent on AI either and you know even the best funds in the world they still have their analysts double-checking the information and they have other sources of data and so that's what you need to do even if from from a retail perspective and and that's what we try to provide is we have different kinds of users right we have researchers and they want to understand data we have people like Goldman Sachs they just want that number in the direction and then we have people that are skeptical about the government so going back to the question it's you're probably not going to be able to be better than AI but you still have your own take AI unless you train your agent like X sentiment right going on Twitter if you're in crypto is so critical because no matter how stupid sometimes it is stupid makes you money like that's the reality in crypto. So >> yeah. >> Yeah. My my body is we're I'm not uh how do you say I'm not skeptical about uh AI also automation in trading you know making their their own decision making and uh performing trades for ourselves. Definitely this is the future. I can see it that uh I I definitely think that uh human curiosity uh human judgment uh and human criticism it's uh in small amounts even using full automation can be uh beneficial and can be healthy. uh this doesn't mean that um um the the conventional trading is going to to be uh but let let's face it even nowadays if you're not using AI in your jailing trading strategies in your analysis you are out of the game it's as simple as that >> and uh definitely we're going uh going there full automation when hard to And the next question is somehow connected. What is better choice? Pay to companies to sell data and signals or second create my own tool to track and do research? >> It's my favorite question. >> I'll keep it I mean I'll keep it very quick. I'm extremely biased because I am a translation. But I will say this in the audience is as I said we build Truman our proprietary theory macro agent and it's traded on our 500 plus real-time indexes. So if you go to Truman and you ask questions about the US economy the correlations I can confidently say that you'll probably be ahead of 99.9% of people on the planet because we forecasted the inflation literally two weeks before the government and we have a 99.3% historical accuracy that you can check because I'm not inventing the number. It's a proven fact. So >> Yep. Yep. current I I don't disagree upon you on on a lot of the points. Uh AI is definitely smarter than us. He's going to be smarter than us. He's going to analyze data stream faster than us, more accurate than us using different uh you know for example hybrid hybrid methods. I definitely on that train. I think that AI is going to overwhelm most of our of our decision making process. And uh on the question uh if uh it is better to for example get signals or indicators from a provider or be build it yourself. I think that both ways are not wrong. If you are capable of uh contextualizing information using AI or derive some uh usefulness of that information using AI, go for it. Why not? But if you for example um a serious trader or a serious analyst and uh uh you a lots of money depends on you, I definitely leave this to the professional guys that are producing uh you know uh products like uh Jordan and myself that can you know can uh be beneficial even for for a payment. I think that is uh worth it definitely. >> Thank you very much. Thank you. >> Thank you very much. Thank you to the audience for staying uh this late. >> Thank you. [music]