267Certvas/african-company-fundamentals

African Company Fundamentals (AF-FUND)

Helps Africa-mandate PE funds and frontier research desks screen and benchmark African listed companies by providing point-in-time, provenance-documented fundamentals so they can underwrite positions without building an extraction team.

License-clean and point-in-time — built for research desks and for retrieval/agent pipelines that need grounding data they can cite.

Why this dataset

Retrieval and agents are only as trustworthy as the data under them. Every value here carries its source_ref (document + page), extraction_confidence, license_class, and observed_at. Published rows contain no natural-person data.

What's inside

• —Income-statement, balance-sheet and per-share line items, normalized to actual units and reporting currency.
• —Keyed by dataco_id, period_end, period_type (FY / H1 / Qn) — point-in-time, so a restatement is a new row, not an overwrite.

Fields

See DICTIONARY.md in the files. Core columns: dataco_id, period_end, period_type, statement, line_item, value_reported, currency_reported, extraction_confidence, source_ref, license_class, observed_at.

Entity identifier — join on dataco_id

Every row is keyed by `dataco_id`, the Certvas ID — a free, open, publicly-resolvable entity identifier (CC0). Join on `dataco_id` to combine this dataset with tenders, macro, or your own data, and resolve any ID for free at https://certvas.com/id/{dataco_id}.

Entities keyed by Certvas ID — free, open identifier: https://certvas.com/certvas-id.html

Use cases

• —Frontier screening — filter revenue, profitability and leverage across the covered universe in one query instead of 40 PDF reads.
• —Benchmarking — compare issuers on as-reported, point-in-time figures.
• —Agent grounding — every value is citable back to its filing (source_ref).

Coverage (proof strip)

This is a free sample (100 rows, full schema). The full product spans 220 issuer-periods and 21,497 facts across 84 issuers in 8 markets, including Safaricom, Dangote Cement, Dangote Sugar, NASCON and Naspers — and grows each release. Published facts carry ≥0.85 extraction confidence; measured coverage/freshness/accuracy ship in metrics.json; per-source licensing in PROVENANCE.md.

Usage

from datasets import load_dataset
ds = load_dataset("267Certvas/african-company-fundamentals")

or read data/gold_fundamentals.parquet directly with pandas/DuckDB.

Provenance & license

PROVENANCE.md lists every source and its license basis; LICENSE.txt covers permitted use (commercial use per the delivery agreement; attribution per each source; no natural-person data). This is a free 100-row sample. Full history, REST API and an MCP server (values returned with provenance) are available in the paid tiers.

Trust & method: https://dataco-site.pages.dev/trust.html · contact mandasetheo@gmail.com.