Pathfinder

The Problem

When a procurement team needs to source a specific part — say a 10" trunnion-mounted ball valve in Class 600 with NACE compliance — they rely on institutional memory, outdated catalogs, or hours of manual research. The knowledge of which manufacturer makes the best fit for a given spec lives in the heads of experienced buyers, not in searchable systems.

The Solution

Pathfinder uses AI with deep industrial domain knowledge to analyze a part specification and return ranked supplier recommendations. It understands pressure classes, material grades, API standards, and manufacturer product lines — the kind of knowledge that typically takes years to accumulate.

Architecture

Browser (paste part spec)
  |
  v
FastAPI Backend (Python)
  |-- Validate: input length, content
  |
  v
LLM API (Domain-Aware Matching)
  |-- System prompt with industrial supplier knowledge
  |-- Structured JSON output with ranking
  |-- Response validation
  |
  v
Browser (render supplier cards + spec analysis)

About

Built by Andres Tobacia — Industrial Engineer with 17+ years in manufacturing, supply chain, and space exploration.

No data is stored. All processing is stateless. AI processing powered by a large language model API. No training is performed on input data.