We extract financial and operating statistics straight from filings with AI, compute valuation multiples and fundamental upside, and run the best strategies as transparent agents you can follow live — across 350+ issuers in markets others barely cover.
Revenue, EBITDA, production, reserves — pulled straight from filing PDFs (SEC, IFRS, KASE, IDX) with LLMs and OCR, no manual entry.
A forward fundamental-potential score on every name, plus we turn today's commodity spot prices into forward EBITDA — so cyclical earnings power shows up before it lands in the financials.
EV/EBITDA, P/E, ROE, net-debt dynamics and fundamental upside — computed automatically and refreshed daily.
For producers we reprice revenue and cost by each commodity's live spot move and revert to the name's own through-cycle EV/EBITDA — with a reserves-life haircut — to surface mispriced cyclicals.
Kazakhstan, Uzbekistan, Indonesia, Mexico, the Philippines — frontier markets where data is scarce everywhere else. We cover them.
Rules-based model portfolios on value, quality and commodity upside — backtested against local benchmarks (results below).
Rules-based agents on our AI-extracted data: validated in backtest, now we run the recommended book live (paper) and mark it to market daily.
Backtest track record
Vectorized equal-weight backtests, monthly rebalance, no costs. Commodity and bank replicas carry survivorship/liquidity bias — treat as a mechanical floor, not the live fund. Full methodology →
9 markets · 350+ companies — click any to open its overview
Or jump straight to any company using the search box at the top.
ai_budget_note