Mike Brusov

The through-line

The single idea Brusov returns to is that neither crowds nor algorithms are sufficient on their own, and that the value sits in the join. Cindicator, as he describes it, exists to build "hybrid intelligence" technology: an app where participants answer financial challenges and try to predict future price movements or market events, after which the collected crowd data is run through data science and machine learning models to produce indicators for traders and investors [1]. His framing is deliberately symmetrical, "it's all about" combining "human and machines" so that two different approaches sit inside one simple product, and he treats that combination as the company's defining asset rather than any single model or dataset [1]. The stated goal is practical rather than theoretical: to create effective instruments that help investors and traders make more relevant decisions [1].

That preoccupation has a personal history behind it. Brusov describes roughly nine years of launching startups and technology companies, beginning with a big data company that built algorithms to analyse social media, collecting mentions and messages from blogs, forums and social networks to create instruments for analysts and marketing managers [1]. He encountered the Cindicator idea four years before the conversation, following a productive exchange with a professor in New York connected to the social experiment on superforecasting, and about two years later the company decided to move to a tokenised model and issue its own token on Ethereum [1]. He sees his own data background and his co-founders' data science and machine learning expertise as a natural fit for the problem [1].

On combining human and artificial intelligence

Brusov is precise about where the differentiation lies. Plenty of companies, he says, already use crowd intelligence to solve problems, and plenty use artificial intelligence on data from traders; the unique difference in his account is that Cindicator combines both types of intelligence in one system [1]. The mechanics he describes run in sequence: collect predictions from the crowd, then apply the data science layer, which he says uses an ensemble of different models including neural networks, combined into a single answer so that the most relevant model is selected for each particular case and each particular prediction [1]. The final indicator is then distributed through the web interface to customers [1].

On what an indicator is actually for

He is careful not to oversell prediction. Financial markets, in his framing, are systems of uncertainty where perfect prediction is impossible; what the platform generates is additional data to help professional players make more profitable trades [1]. Responsibility for outcomes therefore sits partly with the user: "it's very important how you come you implement our data our indicators in your trades" [1]. He backs this with observation rather than theory, noting more than a thousand active customers including traders and fund managers, watched daily, all of them using different strategies according to their own risk profiles and risk management [1]. The indicator is an input to a final decision, not a substitute for one.

On black swans and forecaster clustering

The most interesting cases in the data, by his account, come from segmenting the crowd rather than averaging it. He describes identifying a cluster of forecasters "who have like a very good statistic and chance to predict unpredictable events", grouping them together and drawing on them specifically when the question concerns rare events [1]. He cites this method producing a correct call on a market direction where the majority of people and analysts did not reach the same conclusion [1]. The implication running underneath is that a crowd is not one signal but several, and that knowing which sub-crowd to listen to for which class of question is part of the technology.

On why finance first, and what comes after

Brusov treats financial markets as a starting point chosen for speed of learning rather than as the end state. Because the platform can ask twenty questions a day about different tokens, it can collect enough data points quickly to train networks and generate relevant products [1]. He is explicit that this is "only a beginning", with plans to apply the same technology to business intelligence, business analytics, politics and marketing [1]. The near-term priority he names is more crowd and more data: growing the number of participants and data points so that the resulting products become more valuable [1].

On token sale indicators and full stack analysis

The concrete product ambition he sets out is a new class of indicators aimed at token sales, produced before an investor decides whether to participate [1]. Crucially, he does not want these limited to price prediction. The intention is to analyse the technology and the team as well, using the analyst crowd on the platform to answer assigned tasks, then connecting those human assessments with the AI models into what he calls full stack solutions [1]. It is the same hybrid pattern applied to a qualitative judgement rather than a numeric one.

On competitors, partners and transparency

Asked about rivals, Brusov says "we don't have a direct competitors", attributing this to a specific and unique business model, while acknowledging other companies working on crowd and artificial intelligence approaches for traders, some of them in the blockchain space [1]. He declines to frame them as competitors, preferring to see them as potential partners in developing a shared market intelligence layer, and he takes a broadly welcoming view of rivalry: "competition is one of them like most effective like factor to develop stronger companies stronger products", a driver for entrepreneurs generally [1]. That openness extends to the company's own record. All data and previous predictions are published on the website, which he considers important both for credibility and for engaging additional customers [1]. The same applies to planning: "we have a great tradition to share publicly our product map our business plan with our community", with a detailed roadmap posted on the blog for community members to question and respond to [1].

On the team

He describes a team of fifteen at the time, most of them developers and data scientists, with a strong mathematics and data science bench that includes winners of data science competitions, and he names this concentration of talent as one of the company's main advantages [1].

Takeaways

• The differentiator is the combination itself: many companies use crowd intelligence and many use AI on trader data, but Brusov positions Cindicator as combining both types of intelligence in a single system [1].
• Indicators are decision support under uncertainty, not forecasts to be followed blindly; "it's very important how you come you implement our data our indicators in your trades" [1].
• Segmenting the crowd matters more than averaging it: a dedicated cluster of forecasters good at predicting unpredictable events is used specifically for rare-event questions, and produced a call the majority of analysts missed [1].
• Finance was chosen as the first domain because twenty questions a day generate training data fast; business intelligence, business analytics, politics and marketing are named as later targets [1].
• Publishing all past predictions and the full product roadmap is treated as a customer acquisition and community mechanism, not just disclosure [1].
• Rival crowd-and-AI projects are framed as prospective partners in building a market intelligence layer, with competition seen as a driver of stronger companies and products [1].
• The next product step is token sale indicators that assess technology and team as well as price, combining assigned analyst tasks with AI models into full stack solutions [1].