NYC Job Match

Inspiration

We wanted job search to understand both a candidate’s skills and their career goals. An experienced engineer may want an entry-level opportunity in a new field. Their stated preferences should guide the search.

What it does

NYC Job Match lets users upload a resume, describe their desired role, and find relevant software engineering jobs. Our demo includes 162 postings from 15 employers and NYC Government.

Results show original requirements, published salary information, and AI explanations supported by verified quotes.

How we built it

Mistral converts resume information and preferences into a structured search plan and explains retrieved matches.

Elasticsearch performs full-text retrieval with salary and experience-level filters. Snapshot identifiers exclude outdated postings, while aggregations summarize employers and compensation.

Python powers the backend. HTML, CSS, and JavaScript provide the interface. SQLite saves resumes, application history, subscriptions, and cached AI responses.

Challenges and lessons

Separating requested roles from resume history was a key challenge. We learned to enforce preferences as search constraints and validate AI citations against original postings.

What’s next

We plan to expand employer coverage and add email alerts. The current app supports three-hour refreshes and in-app subscription alerts while running. Our Greenhouse prefilling prototype supports Datadog and Figma; users complete final submission themselves.

Built With

• chatgpt
• elasticsearch
• mistral