Data you can cite.
Certvas helps frontier research desks, market-entry teams and DFI analysts make defensible calls in African & emerging markets by structuring company fundamentals, tender flows and macro context — every value carrying its source document, license class and observation time — so they can screen, size and monitor without building an extraction team.
Start with a free, fully-provenanced sample — validate the citations before you buy.
3,511+ audited-filing facts · 275,733 tender-flow rows · point-in-time macro across 5 core markets · extraction confidence ≥0.85 on every published fact · no natural-person data · 4 provenance guarantees
| Period | Value | Vintage | Known at |
|---|---|---|---|
| 2025-12-31 | 827.40 | v1 | 2026-06-05 |
| 2024-12-31 | 743.19 | v1 | 2026-06-05 |
| 2023-12-31 | 672.53 | v1 | 2024-05-13 |
| 2022-12-31 | 607.20 | v1 | 2023-05-09 |
Three data products, all live.
One schema discipline, one provenance model — across fundamentals, procurement and macro.
Company fundamentals
Multi-year income-statement, balance-sheet and per-share line items for listed African issuers, extracted from published filings and normalized to actual units and reporting currency.
Use it to: screen & benchmark issuers on as-reported figures in one query.
Dictionary, coverage & sample →Public procurement
Open-Contracting (OCDS) tender, award and party records across Zambia, Nigeria and South Africa, with supplier, buyer and monthly analytics. Natural-person parties excluded from every published table.
Use it to: size category demand and identify suppliers before competitors do.
Dictionary, coverage & sample →Macro indicators
Point-in-time macro — daily FX plus monthly CPI (index + YoY) and FX for five core markets, as_of-queryable with no look-ahead — over an annual World Bank context layer, every value cited to its series.
Use it to: run clean, no-look-ahead backtests and join macro context to company and tender signals on one spine.
Dictionary, coverage & sample →The same gold tables, delivered three ways.
Every surface wraps each response in a provenance envelope your model can quote back to a user.
Files
Parquet + CSV with provenance, dictionary and metrics bundled in a dated, citable release.
REST API
Key-authenticated, rate-limited, OpenAPI-described — every response carries its license notice. Also listed on the RapidAPI marketplace.
MCP server
Agent-native tools that return values with their source and license, so your model can cite them. Published to the official MCP registry, so agents can discover it.
Built for the whole desk — not just one seat.
One license-clean, point-in-time core, delivered so every team can plug it in. See how Certvas clears a fund’s bar →
Quant & systematic
Point-in-time panels with as_of and no look-ahead, stable IDs to join your security master, bulk Parquet. Backtest without rebuilding the joins.
AI, RAG & agents
Every value returns with its source and license via the API and MCP server — grounding data your model can cite, that survives a compliance review.
Research & discretionary
Demand signals, factors and macro context with citations — defensible country and sector views in an afternoon, not a data-engineering sprint.
Engineering & data
pip install certvas, one schema across REST / MCP / Snowflake / files, provenance on every row. Integrate in an afternoon.
Start with a free sample.
Every tier — including the free sample — ships with the full schema, provenance annex and measured metrics. Validate the citations end-to-end, then scale to the full history.