חני פרוינד

I built and run JobMatch AI, a SaaS product that pairs a Chrome extension with an async Python and FastAPI backend. It creates a daily deck of matching jobs and tailors a CV to each job, and is used by close to 1,000 job seekers.

JobMatch AI collects listings from company career boards and LinkedIn, then normalizes, deduplicates, and embeds them in pgvector. Per-user vector search feeds parallel Claude calls, while deterministic code calculates an explainable match score.

At Matrix IT, outsourced to Infinya (Hadera Paper), I built Java modules with Hibernate and PostgreSQL for production systems. I redesigned a plant-data task, cutting its resource consumption by about 70%, and improved PostgreSQL query execution time by up to 80% across production modules.

I also built Kesher Tov, a Hebrew RTL community job-help platform using Next.js, TypeScript, and Supabase, with unit and end-to-end tests in CI. Earlier, I lectured in mathematics and computer science and built a classroom management platform with React, Node.js, and MySQL.