Daniel Shea and Nasrudin Salim on Chatoyance

Author(s): Scott Douglas Jacobsen

Publication (Outlet/Website): The Good Men Project

Publication Date (yyyy/mm/dd): 2025/01/14

Daniel Shea, M.Sc. is the founder and CEO of Chatoyance. Shea possesses a Master’s degree in Computer Science from the University of New Hampshire, with several years of industry experience in software engineering. He has published freelance articles on foreign exchange market strategy analysis and has published software analyzing fractals in the foreign exchange markets. Leveraging his experience with software design and financial markets, he started Chatoyance with the intent of transforming the way independent investors approach the foreign exchange market. 

Nasrudin Salim is the Co-Founder, COO and CTO of Chatoyance. He has worked in the financial trading and banking industry specializing in machine learning and previously headed the ML operations team in DBS Bank, led AI architecture in OCBC Bank, the 2 of the largest banks in Singapore and Asia and was VP of Engineering in Almanak which uses AI agents for on-chain trading in web3. His specialty is in building machine learning and AI systems at scale and also in real-time processing.

Scott Douglas Jacobsen: When did you two meet?

Daniel Shea: We first met in 2012 in a high IQ society called Torr. Nasrudin had posted an internal message to the group about his recent experiences trading on the foreign exchange market, and I followed up with my own. We discussed more offline, then started working on independent trading projects with each other. One such project was a platform that allowed us to automatically mirror each other’s trades via a central server with which our separate trading platforms would communicate. We then realized we could scale this up to a wider audience, and Chatoyance was born.

Nasrudin Salim: In 2012, I was an 18 year old back then, having started trading at the age 14 with my parent’s money. I did a bit of bitcoin and forex and found success during a time when the market was not as volatile and full of trading agents and bots like today. I posted some insights into a high IQ society called Torr which had a minimum IQ requirement to join at 146, percentile at the 99.87th. Dan replied to some of my posts and we realized we both approached trading from a systems engineering perspective. At first we did simple trading projects, and then later we came to the idea of building a sort of trade sharing collective. Dan did most of the work initially as I didn’t know how to code much back then but grew rapidly later. We started building custom integrations to mirror each other’s trades on the popular platform MetaTrader 4. Then eventually it was about mirroring everyone in a group, not just one-way but bidirectional as many-to-many communication.

Jacobsen: What was the origin of the idea for Chatoyance?

Shea: Chatoyance initially started as a social trading platform which, as mentioned, was itself started as a means for us to share trades in real-time. This gradually evolved into a platform that generated trading strategies based on predefined characteristics using genetic programming. Though these two services would seem quite distinct, there are some core similarities, chief among them being the idea that many strategies operating in parallel outweigh a lone strategy over time and that there is a constant need to reevaluate and cycle out strategies as market conditions evolve.

Nasrudin Salim:  Early on, we thought, “why limit these mirrored trades to just us?” Both of us were layering signals, blending sentiment and quant metrics. The strategy seemed scalable and liquidity was deep. The original concept was basically a distributed, real-time signal exchange. It was like a sandbox where multiple strategies or traders could compete, evolve, and reinforce each other. As the system matured, we introduced genetic programming to shape custom strategies on the fly. So, from the start, the seed idea was that multiple concurrent approaches can minimize single-strategy fragility. That’s how Chatoyance was born.

Jacobsen: How has the business and technology, and software, landscape for Chatoyance’s focus changed in the last ten years?

Shea: There is certainly more competition in this space now than there was one decade ago. This is likely due to the lower barrier to entry and a hype cycle when it comes to AI. Some of the core tech has changed over time to reflect advances in the field. But another change has been the interest in different asset classes over time. Our software is designed to accommodate currency pairs, equities, commodities, cryptocurrencies, and more, but interest from clients has shifted over the years. Forex was the initial interest one decade ago. These days, equities and cryptocurrencies are asked about more regularly.

Nasrudin Salim: The stack is radically different. A decade ago, market data pipelines were heavier and less real-time. Now, I have a cheap feed of tick-level crypto, forex, equities and also options data and can run complex ML models, even LLMs directly on live streams. Cloud infra matured, open-source AI toolkits exploded, and more competition due to now a lower barrier to entry. We’ve seen forex become less sexy and crypto become standard for high-risk plays. I had to ensure the underlying architecture scales to new asset classes fluidly. We’re definitely dealing with a more fragmented but also more flexible ecosystem.

Jacobsen: How is machine learning and AI built into the business?

Shea: The core product that we offer to clients is a service that automates the construction of trading strategies based on current market conditions. Additional tiers involve full portfolios, that is to say many strategies of different trading styles or risk tolerances per the desires of the client, and strategies that evolve as market conditions change over time, owing to the fact that any strategy which works in the short term is unlikely to hold for long. This is ultimately done by leveraging AI. That is said with the full acknowledgement that the term “AI” can be quite loaded and overused these days, often used to placate certain audiences. Despite the current implications of the term, there is indeed no better term to describe what is being done. With that said, just about anyone could develop an application that outputs strategies by the end of a weekend-long hackathon. The breadth of technical indicators used, entry and exit strategy logic employed, optimization criteria supported, money management strategies considered, and robust filtering logic included all coalesces to form a more comprehensive offering than competing organizations.

Nasrudin Salim: We apply ML from the ground up. Every piece of the puzzle from market microstructure to anomaly detection, dynamic portfolio rebalancing. We mix between simple algorithms, genetic optimization to traditional machine learning, then to reinforcement learning and now LLMs. The key is continual learning. Strategies adapt as new conditions emerge and so do the humans who now build how these strategies are going to adapt. Like including meta-learning concepts, model ensembles, and reinforcement signals. The result is that you’re not stuck with stale logic. It morphs as volatility regimes shift or as new liquidity venues pop up.

Jacobsen: How does Chatoyance build more social trading into the trader networks?

Shea: The first iteration of Chatoyance was a more social experience. The idea was that there would be different trading rooms, and members of these rooms would automatically copy each other’s trades through our software. There would be safeguards in place, such as the option of enabling private rooms, muting certain traders so they could only receive trades but not contribute any to the group themselves, and so on. The idea was that, if you had a room of traders each interacting with the markets, the collective gains would outweigh the collective losses, resulting in everyone benefiting from the participants’ engagement.

The business model was that users registered with an affiliated broker, and thus commission was collected on each trade. Since a single trade was replicated for each user in a trading room, this meant a single action from a user could result in wider commissions due to each member simultaneously opening or closing the trade.

In practice, this was not quite the case. Often, people would join trading rooms and wait for others to make the first move. Those who were more experienced did not feel a motivation to contribute trades without some clearer incentive. Some ideas, such as profit sharing on commission, were proposed, but ultimately, if someone is skilled at swing trading the markets, they are more likely to go into fund management themselves than potentially risk it all on some other member running a huge drawdown.

So the idea was ultimately scrapped after several months. However, the idea of many traders bringing their own strategies to a collective single trading room has a spiritual line to our later concept of automated strategy generation with distinct trading personalities, together constructing an automated portfolio.

Nasrudin Salim: We learned that simple social mirroring wasn’t sticky. Traders either lurk or they just want someone’s edge without giving their own. So instead, we integrated the “social” element into a collaborative network of AI-driven strategy modules. Each “node” in the network is like a trader with a personality. From maybe momentum-focused, or mean-reversion-heavy, and they collaborate by sharing signals and outcomes. It’s less about people copying each other and more about these agent-like strategies feeding into each other’s learning loops, evolving collectively to handle shifting regimes. It’s social trading, but via synthetic participant strategies rather than pure human interaction.

Jacobsen: How do you do risk management?

Shea: Risk management is particular to the client, but there are many levers to pull when assessing one’s risk tolerance. Risk management can range from high-level goals, such as drawdown thresholds and Sharpe ratio targets, to finer-grained details such as exit strategies, money management strategies, partial entries and exits, and more. Many times, people will state that they want a high-risk high-reward strategy, but suddenly get cold feet at the first sight of what that risk entails. There is an element of getting to the heart of one’s true risk tolerance before crafting a template that generates appropriate strategies.

Nasrudin Salim:  Risk management is programmatic and multi-layered. For crypto, for example, I might impose real-time volatility-adjusted position limits. For a more traditional asset, we might weigh by a blend of sector correlation risk and liquidity depth. The user sets broad tolerances like max drawdown or desired sorting ratio. From there, the ML system translates that into execution-level heuristics. The idea is we fuse top-down constraints with bottom-up adaptive strategies.

Jacobsen: How do fractals play into financial markets?

Shea: Fractals are one indicator among many that are baked into the product. The algorithm may use fractals depending on market conditions, but may not. The interest in fractals in particular comes from an old technical indicator that was published to the MQL Marketplace (https://www.mql5.com/en/market/product/4131). However, in the current iteration of the product, it is not highlighted any more prominently than additional indicators, ranging from the standard basket (ADX, ATR, CCI, EMA, MACD, RSI, etc.) to the more esoteric (candlestick patterns, Fibonacci retracements, Elliott Waves, etc.) depending on the interests of the client.

Nasrudin Salim: Maybe fractal-based signals matter in certain trending conditions or where micro-structure has repeating patterns. if the system thinks fractals add incremental predictive power given current conditions, it’ll use them. As one of the architects of Chatoyance, I add it as just another tool that our systems could use, and the choice is autonomous. If not, it won’t. We never rely on a single tool. Everything competes on a data-driven meritocracy.

Jacobsen: What are the challenges facing technology-driven financial companies?

Shea: At least from the conversations I have with others in this space, I notice that there is often an overreliance on technical indicators at the cost of fundamentals. This makes sense from a programmatic perspective as engineers can readily integrate these into their models. With that said, the fusion of technicals and fundamentals is necessary to arrive at a more holistic view of the market, all of which serves to only improve the outputs of the algorithm.

Nasrudin Salim: One of the big ones is bridging the gap between what’s quantifiable and what’s real. Pure technical systems might ignore underlying credit conditions, macro news, or liquidity crises until it’s too late. Also, data noise, market manipulation, and wild regulatory shifts can break your models. It’s crucial to design adaptive frameworks that don’t assume static conditions. We’re constantly at war with overfitting and model drift. Especially in cryptocurrency where a lot of the movements originate from insider activity and information found in web3 ‘Cabals’ that exists as Telegram group chats, which can only be joined through connections or NFT purchases.

Jacobsen: What are the guiding principles of Chatoyance?

Shea: It is deceptively simple to say that one’s financial goals are just to “make lots of money.” As discussed earlier, people may feel confident moving forward with a high-risk high-reward strategy at first, only to recoil at the first drop. This isn’t entirely unexpected; after all, a safer market experience would be to invest in a set-and-forget whole market ETF. To pursue these strategies is to expect higher reward at the cost of higher risk. However, even in this more narrow range of higher risk tolerance, there is a wide window of consideration and opportunity. We ultimately aim to reconcile this risk-reward trade-off on a per-client basis and arrive at a portfolio that doesn’t fail to impress.

Nasrudin Salim: We want to democratize robust strategy generation. It’s not just “make money fast.” it’s “craft a strategy that aligns with your true risk appetite and thrives under evolving conditions.” We want to give clients a toolkit that doesn’t lock them into a fixed view of markets. Instead, we shape a pipeline that constantly checks itself like adjusting parameters, evaluating signals, pruning weak strategies, doubling down on robust ones.

Jacobsen: Thank you for the opportunity and your time.

Shea: Thank you for giving us the opportunity to highlight what we have built! This space moves slow and then fast all at once. The journey has been edifying, humbling, and exhilarating. We have many years behind us and are looking forward to many more.

Nasrudin Salim: Happy to share what we’re up to. It’s been good to lay it all out.

Chatoyance Pte. Ltd. (“Chatoyance”) The materials and data contained on this website and any related mobile application are for information only and shall in no event be construed as an offer to purchase or sell, or the solicitation of an offer to purchase or sell, any securities in any jurisdiction. Chatoyance does not make any representation, undertaking, warranty, or guarantee as to the timeliness, completeness, correctness, reliability, or accuracy of the materials and data herein. Certain statements made on this Site may not be based on historical information or facts and may be “forward looking statements”. Actual investment results may differ materially from these forward-looking statements due to a number of factors, including future changes or developments in the business of a company featured on this Site or other political, economic, legal, and social conditions. All opinions, forecasts, or estimates expressed herein are subject to change without prior notice. Chatoyance and its affiliates accept no liability or responsibility whatsoever for any direct or consequential loss and/or damages arising out of or in relation to any use of opinions, forecasts, materials, and data contained herein or otherwise arising in connection therewith.

Last updated May 3, 2025. These terms govern all In Sight Publishing content—past, present, and future—and supersede any prior notices. In Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons BY‑NC‑ND 4.0; © In Sight Publishing by Scott Douglas Jacobsen 2012–Present. All trademarks, performances, databases & branding are owned by their rights holders; no use without permission. Unauthorized copying, modification, framing or public communication is prohibited. External links are not endorsed. Cookies & tracking require consent, and data processing complies with PIPEDA & GDPR; no data from children < 13 (COPPA). Content meets WCAG 2.1 AA under the Accessible Canada Act & is preserved in open archival formats with backups. Excerpts & links require full credit & hyperlink; limited quoting under fair-dealing & fair-use. All content is informational; no liability for errors or omissions: Feedback welcome, and verified errors corrected promptly. For permissions or DMCA notices, email: scott.jacobsen2025@gmail.com. Site use is governed by BC laws; content is “as‑is,” liability limited, users indemnify us; moral, performers’ & database sui generis rights reserved.