Job Research
AI job-search product at a startup · 2025 to 2026
A conversational job-search assistant for frontline workers, the people who stock warehouses, run registers and care for patients. I was the technical lead of that app, and I built the pipeline that filled its jobs database alone.
- The problem
- Frontline workers need job-search tools that fit their careers.
- What I did
- I led the app and built its data pipeline alone.
- The result
- A product that finds the jobs that fit your profile and helps you grow.
Frontline workers search for jobs with tools built for someone else
The people who stock warehouses, run registers and care for patients look for work with tools designed for corporate profiles: keyword filters, resumes, dashboards. We set out to give them a conversation instead. Ask a plain question and one screen answers it: an AI summary, real listings pulled from live search, the sources it used, and what to ask next.
I led the app and built its data pipeline
I was the app's technical lead from its first commit. I also built its data pipeline on my own, keeping the jobs database supplied with listings. AI powered the product; we wrote the code by hand with help from a coding assistant.
Parallel searches bring relevant jobs into one answer.
A person asks a question about their career. The model checks whether it is about jobs and uses the conversation to turn it into a search query. The app then searches the sources they selected: the web, live job listings and videos. It combines those results with the profile from their uploaded resume to suggest relevant jobs and next steps.
I turned raw listings into searchable job pages.
I built the pipeline that supplied the assistant with jobs. It imported listings from several feeds, added company information and used AI to standardize titles, pay and locations. Each listing got a public page that search engines could index. The stages ran as separate background jobs in batches, so a failed listing did not stop the whole run.
Fast models for search steps, strong ones for answers.
Each step had a specific task and a defined response format. I chose faster models for preparing queries and summarizing search results, and stronger models for reasoning and writing the final answer. The aim was to balance useful answers with a reasonable wait. One step also had a fixed limit on its input text.
Two apps, one product, and the patterns I still use
I shipped two apps that made one product. The research app answered in natural language and put an AI summary, real listings, their sources and the next questions to ask on one screen. The pipeline behind it filled the jobs and companies database, normalized every listing, and gave each one its own public page. Those pages were what search engines could index.
I can turn a language model into a product feature, with a clear task, prompt, validated response format and suitable model for each step.