AI Component Data Enrichment Workflow
A deployed portfolio web app that enriches component descriptions from tabular input, demonstrating acty.dev's AI Engineering and data enrichment work.
bomEnricher is acty.dev’s reference implementation for AI-assisted component data enrichment on spreadsheet input.
Proof assets and links
Problem
Procurement and engineering teams keep component lists in spreadsheets where descriptions are inconsistent, abbreviated, or incomplete, which slows search, comparison, and ordering.
Context
bomEnricher is a portfolio project, not a public SaaS product or a client deployment. It shows the pattern acty.dev uses for AI-assisted data enrichment on spreadsheet data.
Inputs
- A bill-of-materials or component spreadsheet
- Component identifiers or partial descriptions
Technical approach
The app ingests a spreadsheet of components and applies AI-assisted enrichment to normalize and complete the component descriptions, then returns an enriched table for download. The deployed instance demonstrates the ingest, enrich, and export flow.
Stack
- TypeScript / JavaScript web app
- Spreadsheet (Excel) processing
- AI/LLM-assisted enrichment
- Deployed on Replit
What was built
A web interface that accepts a component spreadsheet, runs AI-assisted description enrichment, and produces an enriched, downloadable table. It is the reference implementation behind acty.dev's data enrichment service.
Deployment
Deployed as a portfolio web app at descript.replit.app (verified reachable 2026-06-03; the Replit instance cold-starts on the first request).
What can be reused
- Spreadsheet ingest and export pipeline
- AI-assisted field normalization and completion
- Human-reviewable enrichment output
Boundaries
- Portfolio proof, not a public SaaS product
- Client and private data are not included
- Do not claim measured labor savings without a separate sourced calculation
