consulting Active

Bid & Investment Screening for an Infrastructure Group

A full priced tender package produced in 3 days for a six-country infrastructure group, plus an investment screen that caught a falsified growth claim before money moved.
Role
Designed & built solo
Stack
Claude, Python, Parametric estimation, Financial analysis
Outcome
Full priced bid in 3 days, against a 4 to 6 week team baseline

Problem

A six-country infrastructure group had a major facilities tender on a hard external deadline. Producing a serious bid for work like this normally takes a team 4 to 6 weeks: technical compliance mapping, risk analysis, local-content planning, and a fully priced commercial offer. The window they had was days, and I was working from limited context, without full access to their internal data.

Separately, the same group was weighing an investment in a company whose pitch rested on a claim of consistent growth.

Approach

For the tender, I ran the whole package through an AI-assisted build. Parametric estimation from a bill of materials and local labor rates produced the cost base. On top of that came the layers a review committee actually scores: a clause-by-clause compliance map against the tender requirements, a 40-item risk register, a 74% local-content plan, and a priced commercial template ready for submission.

For the investment screen, I did what a diligence team does and let the numbers speak. I rebuilt the target’s revenue picture from their own documents instead of accepting the summary slide.

Outcome

The full priced tender package, a bid recommendation of USD 7.73M, went out within the deadline. Work the group normally scopes in 4 to 6 weeks was produced in 3 days, from limited context. With full access to their data, the cycle gets shorter still.

The investment screen caught the story breaking. The “consistent growth” claim was false: revenue was actually down 23% year on year. That finding reached the decision-makers before any money moved.

Both deliverables are real work product for a real organization. The client is not named here by agreement; the outcomes are the point.

What I’d Do Differently

I priced the tender from public benchmarks and inference where internal data should have been. It held up, the estimate landed close to the group’s own numbers, but that was earned the hard way. Next engagement starts with a data-access conversation before the clock starts, not after.

Key Metrics

Metric Result Benchmark
Tender package Full priced bid in 3 days, against a 4 to 6 week team baseline n/a
Package depth Compliance map, 40-risk register, 74% local-content plan, priced commercials n/a
Investment screen Caught a falsified growth claim, revenue actually down 23% year on year n/a