Why I stopped trading and started tendering with AI

Description and use

The author replaced trading bots with an AI system for tender analysis and response, boosting bid volume, win rate, and cutting costs.

Problem

Traditional tendering is manual, time‑consuming, and error‑prone, making it infeasible for small businesses to respond to many tenders.

Outcome

The AI pipeline increased bid volume, improved win rate, reduced manual costs, and provided scalability for tender responses.

Workflow

• Parse tender documents with OCR (PyTesseract, pdfminer.six)
• Chunk and summarize text
• Use a fine‑tuned Llama 2 7B for requirement extraction
• Embed requirements and query Pinecone for matching capabilities
• Generate compliant tender response with LLM prompts
• Cross‑check response against original requirements for compliance

Evidence

Traditional methods relied on manual reading and tagging - a process slow and expensive.

Source

I fine‑tuned a model (currently a quantized Llama 2 7B, deployed locally for cost and control) on a dataset of tender documents and manually labeled requirements.

Source

We can now analyze and respond to significantly more tenders.

Source