Tender Getter RSA

Tender Getter RSA Logo

Inspiration

In South Africa, small businesses and "tenderpreneurs" are the lifeblood of job creation and economic growth. Yet, thousands of capable enterprises are locked out of life-changing public and private contract opportunities. The culprit? An overwhelming, complex procurement system.

Sifting through over 600 fragmented municipal, provincial, and state-owned enterprise (SOE) portals is a full-time job. Even when an opportunity is found, a business owner must wade through 100-page Standard Bidding Documents (SBDs) to find hidden compliance traps. A single missing tax PIN, an outdated certified ID, or a slight mismatch in CIDB grading results in instant, cold disqualification.

We built Tender Getter RSA to save ourselves first, and then save our communities. It is an AI-native co-pilot designed to turn bureaucratic lottery into an automated, highly-predictive matching engine.

What it does

Tender Getter RSA operates as an end-to-end, local-first procurement accelerator:

Intelligent Ingestion: Monitors, cleans, and aggregates tender notices daily from eTenders, municipal bulletins, Eskom, Transnet, and SANRAL.
The CSD/CIDB Gatekeeper: Instantly screens and filters tenders against a company's unique profile (including CSD registration status, BBBEE level, CIDB grading limits, and geographical local-content requirements) before they waste time reading.
Gemini Compliance Sieve: Parses sprawling bidding PDFs in seconds, extracting a clear, 10-point mandatory compliance checklist and mapping out the exact functionality scorecard (e.g., scoring requirements for experience and methodology).
BidCraft Proposal Co-Pilot: Drafts highly professional Tender Executive Summaries, custom Project Methodologies, and compliance response packages aligned perfectly with the tender’s evaluation guidelines.

How we built it

We designed Tender Getter RSA to be lean, offline-first, and highly reliable:

Core Backend: Python-first architecture using Pydantic for strict data-schema validation of companies and tenders.
The Brain: Google Gemini (via Google AI Studio / Gemini API) handles heavy-document OCR, compliance extraction, and proposal draft synthesis without look-ahead bias or context bloat.
Data Layer: SQLite for high-speed local matching and caching.
Business Verification: Outbound pipeline utilizing CIPC registration directories and public CSD vendor databases to target and onboard high-intent local companies.

Challenges we faced

The primary challenge is the sheer messiness of South African tender documents. Scanned, skewed, multi-hundred-page PDFs containing legacy tabular formatting are notoriously difficult for standard parsers to understand. We had to design highly defensive prompt-chaining pipelines that segment documents into critical SBD compliance components, keeping token costs low and extraction accuracy high.

What we learned

We learned that the ultimate value is not in creating a longer list of links—businesses are already drowning in email alerts. The true value is in getting the tender. By automating the screening of non-negotiable gates (such as CIDB grade limits and local-presence constraints) and matching them directly to verified company scorecards, we can instantly tell a business owner whether they have a realistic path to winning a contract.

What's next for Tender Getter RSA

Our roadmap is locked into four milestones:

Phase 1: Proof of Concept (POC) Verification: Generating free customized "Tender Opportunity Reports" for local construction, cleaning, and security firms to prove direct value.
Phase 2: Commercial Launch: Rolling out our three-tier pricing model (Zero Upfront Success Fee, Single-Tender Deposits, and R299/month MRR Subscriptions).
Phase 3: Deep Document Integration: Expanding automatic form filling for SARS Tax PIN checks, COIDA Letters of Good Standing, and CIPC structures.
Phase 4: Clones & Expansion: Cloning the underlying matching-and-synthesis architecture to launch sister products: "Job Getter" and "Funding Getter" globally.