EquationDB is a market database and research engine. Ask any market question in plain English or one query language, and get the answer with the evidence, across 70,000 instruments on 70 exchanges, running 24x7.
It replaces the stack of price feeds, overnight batch jobs, notebooks, alerting tools and spreadsheets a desk normally needs, and keeps working after your team goes home.
A purpose-built market database for people and AI agents. Indicators are computed as data arrives, so a whole-market screen returns in milliseconds. Market events are stored with what followed, so studies use only what was known on the day. Every answer shows the exact statement that produced it.
Faster decisions. An event study, a "why did it move" or a scenario is one statement answered in seconds, not days.
Better-tested ideas. Beliefs graded on decades of history by one fixed rule; edges judged after costs.
Research that keeps itself honest. Every night it re-tests its beliefs and scores its own paper calls, misses included.
Wider coverage, one system. Global equities, funds, FX and crypto with fundamentals, transcripts and themes.
| Ask anything | Plain English, one query language or full SQL | Events | 1,515 event types, rarity-scored, with what followed |
| Why it moved | Each driver's share of a move, plus the unexplained part | Backtest | Stops, costs, next-bar fills; same code as live alerts |
| Simulations | "What if": how a target responded every time before | Series | Any price, statement line or flow as one time series |
| Symbol pages | Indicators, levels, events, peers, fundamentals for one stock | Taxonomy & themes | Modern 8-level industry tree; 2,300 themes with indexes |
| Knowledge graph | Suppliers, customers, competitors; shock propagation | Thesis | Belief in plain English → graded verdict; 259 tested theses |
| Knowledge | Learns nightly; makes and scores its own paper calls | Alpha Lab | Which anomalies survive costs, for Gulf and emerging markets |
| Earnings calls | Measured, read by AI and quoted; search across companies | Genome · MarketEye | Every formation in a stock's history; live market map |
| Apps & dashboards | Save any query as an app; link apps on dashboards | Demo library | 1,500+ power queries, explained and ready to run; new daily |
| 13 languages | AI analysis and the interface, Arabic included | Always on | Alerts, scheduled tasks and reports, daily brief, AI agents |
Figures describe the live product in October 2026; run SHOW DATASETS in the app for current counts. Cost figures are illustrative and rest on the assumptions stated. EquationDB reports historical evidence with its sample size; it does not forecast or promise returns.
| Recurring work done for you | Assumed by hand | h / yr |
|---|---|---|
| Post-close desk brief | 2 h × 250 days | 500 |
| Morning screens and alert checks | 1 h × 250 days | 250 |
| Earnings-call review | 100 cos × 4 × 1 h | 400 |
| Re-testing a thesis book | 259 × 4 / yr × 2 h | 2,072 |
| Total | ≈ 1.6 analysts | 3,222 |
≈ $322k a year at $100/h ($200k fully loaded ÷ 2,000 h). Illustrative; substitute your own figures.
The 24x7 comparison. A minimal in-house desk to build and run this capability (2 data engineers + 2 quant analysts at an assumed $200k each, fully loaded) costs about $0.8M a year and covers a 40-hour week, 24% of the week's 168 hours. Matching it around the clock takes 4.2× the people: ≈ 17 FTE, ≈ $3.4M a year.
Excludes data licences, infrastructure, recruiting, management and turnover, all of which raise the in-house figure. EquationDB does not replace analysts' judgment; it removes the plumbing and the overnight grind.
No native spreadsheet add-in; spreadsheets connect via CSV export or the API.
One plan, every feature. No tiers and no feature gates. Usage limits exist only to protect the service, are the same for every firm, and can be raised for yours (for example 20,000 rows per query, 2,000 requests per minute per firm).
Free trial of that same plan for your team. Pricing on request, per firm, against the cost of the work it replaces.