Public 12‑week algo project
Trader and developer Virat Singh began a 12‑week, public algorithmic‑trading project funded with a $10,000 account and open‑source code so others can inspect his build and results. He’s posting trades, code, and progress updates on X so followers can trace his dataset choices and execution decisions in near real time. (x.com)
Virat Singh has started a 12-week public trading build with a $10,000 account, posting the code, trades, and updates in public as he goes. (x.com) Algorithmic trading means using software to turn market data into buy or sell decisions, then sending those orders through a broker automatically or semi-automatically. Singh has framed this project as a live build log rather than a closed fund, with followers able to inspect each change as it happens on X. (x.com) Singh is already known in developer circles for open-source finance projects. His GitHub profile lists New York as his location, and his “ai-hedge-fund” repository had about 54,900 stars and 9,500 forks when it was crawled this week. (github.com 1) (github.com 2) That earlier repository is explicitly labeled a proof of concept and says it is “for educational purposes only” and “not intended for real trading or investment.” The new 12-week challenge moves from simulated decision-making toward a publicly tracked account with real money on the line. (github.com) (x.com) The codebase behind Singh’s older project shows how these systems are usually assembled: separate agents pull market data, score fundamentals, read sentiment, measure risk, and then hand a final decision to a portfolio manager. His repository also lists well-known investor-inspired agents alongside valuation, sentiment, fundamentals, technical analysis, risk, and portfolio roles. (github.com) Public trading journals are common on social media, but most show screenshots after the fact. Singh’s setup is closer to an open notebook: the account size is fixed at $10,000, the timeline is fixed at 12 weeks, and the code is available for other people to audit, copy, or challenge. (x.com) That openness also makes the weak points easier to spot. A public code trail lets other developers see whether a strategy depends on curve-fitting past data, delayed execution, or selective trade reporting, problems that often stay hidden in private backtests. (x.com) (github.com) The next 12 weeks will test whether Singh can turn an audience-facing software project into a transparent trading record. If the posts continue as promised, followers will be able to compare the code, the data inputs, and the account results in near real time. (x.com)
