SimicX
SimicX provides an AI-driven quant research infrastructure that reads published research across languages and disciplines, breaks studies into testable ideas, and fuses compatible concepts into novel trading rules. The platform runs strict walk-forward backtests in a sealed sandbox with net-of-cost execution and deflated performance metrics, delivering code, backtests, and reports sufficient to re-run the work. Institutional deployments run inside the client's environment with proprietary data and risk limits.
- Artificial Intelligence
- Financial Technology
- Data & Analytics
- Software Only
Funding
Raised to date
$120KRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Founder details are not available yet.
Product
Problem
Most published quant research does not survive real-market scrutiny, and once a factor is published, its portfolio returns fall by 58% on average. Research teams cannot read all new studies across languages and disciplines, let alone test them individually, making it costly and slow to separate durable alpha from noise.
Solution
SimicX operates an AI-driven research pipeline that reads published research—working papers, journals, classic texts, broker research, and other sciences—in their original languages, breaks studies into testable ideas, and fuses compatible ideas across unrelated studies into new trading rules. A second AI plays the skeptic, running strict walk-forward backtests in a sealed sandbox where each day sees only that day's real market data, every trade pays costs, and results are deflated for every trial. The platform delivers code, backtests, and reports sufficient to re-run the work, and institutional deployments run inside the client's environment with proprietary data, existing exposures, and risk limits. Every test result flows back into a knowledge base, so the system never forgets and never runs the same test twice.
Target Audience
Primary customers are portfolio managers and investment leaders at asset managers and hedge funds who need scalable, validated alpha research without expanding in-house research teams.
Features
- AI scientists read all published research within days of release, including Chinese-language sources, covering working papers, journals, broker research, and non-finance fields
- Studies are split into individual testable ideas, each linked to the exact source passage, with weak or off-topic work screened out
- Fusion engine combines ideas across studies that never cite each other, enabling novel trading rules no human team could manually discover
- Sealed-sandbox backtesting with walk-forward methodology, net-of-cost execution, and deflated performance metrics to prevent overfitting
- Institutional deployment (AlphaCoder) runs inside the client's environment, incorporating proprietary and vendor data, existing exposures, and risk limits
- Alpha Foundry merges multiple candidate alphas into a single book, removing disguised duplicates and applying inverse-volatility weights within risk budgets
- Alpha Mutation loop retains rejected experiments as memory, enabling each round to start where the last ended