ZAM

Market sizing for founders, with TAM, SAM, and SOM grounded in retrieved benchmarks

Solo project · Solo builder · December 2025 — June 2026

Overview

ZAM is a market sizing tool for founders. A four step wizard collects product, target market, go to market, and competitive context, then produces TAM, SAM, and SOM scenarios with an AI written analysis. I built it solo, all 108 commits.

How it works

Sizing is top down: TAM is potential customers times average contract value, SAM narrows it by geography, segment, and reach, and SOM applies a realistic share target. Assumptions can be adjusted with instant recalculation, and analyses are saved to a portfolio backed by MongoDB.

AI analysis

The report is grounded with retrieval. An in-process BM25 index over industry benchmarks, comparable companies, and the sizing methodology feeds the prompt, and Groq runs Llama 3.3 70B with a smaller Llama model as fallback. In June 2026 I replaced Gemini embeddings with BM25 and moved generation from Gemini to Groq.

Status

Live on Vercel, with a React and Vite frontend, an Express API, and per IP rate limiting on saved analyses.