AutomationIn use
Tender Alignment Engine
Scores public tenders against a company’s capabilities to show which are worth bidding on.
The problem
Reading through daily government tender notices is slow, and bidders can spend money on proposals they were never eligible to win.
Approach
A scheduled pipeline collects listings from Zimbabwe’s public procurement portal and extracts mandatory requirements, financial limits and scope. A company capability profile in YAML is scored against each tender to produce an alignment rating and a list of the specific gaps.
Key decisions
- Deterministic rules decide eligibility; semantic matching only ranks, so compliance results are explainable.
- Each tender gets one of three verdicts: winnable now, winnable after closing named gaps, or out of reach.
- A source-agnostic schema leaves room for other regional procurement portals.
Highlights
- Turns a manual daily scan into an automated report.
- Shows exactly which certifications or documents are missing for a given tender.
Stack
- Python
- FastAPI
- YAML
- Web scraping
- NLP