JobPilot

A Next.js job search copilot that discovers early-career software roles, ranks them against your profile, parses your resume into structured data, and optionally uses an LLM to draft tailored application answers.

What works

• Runs as a Next.js app with API routes for discovery and LLM drafting.
• Parses resume text into a structured profile: experience, education, skills, summary, and contact details.
• Scores jobs against enabled search profiles.
• Applies hard filters for required keywords, excluded keywords, seniority, maximum experience requirement, location, employment type, and salary minimum.
• Keeps profile matching separate from application decisions, so you can review jobs manually.
• Optionally calls an OpenAI-compatible LLM endpoint to generate tailored "why company", AI experience, cover note, match summary, and keyword drafts.
• Discovers jobs from configured public Greenhouse, Lever, and Ashby company boards and ranks them.
• Discovers new-grad and internship roles from the SimplifyJobs GitHub lists, similar to SWEList.
• Optionally searches the broader web from your enabled search profiles through SerpAPI or Google Custom Search, then normalizes and ranks those results with the same matching engine.
• Stores local app state in browser localStorage during development.

Run locally

npm install
npm run dev

Then open http://localhost:3000.

Use

1. Open the Profile tab and paste resume text.
2. Parse the resume into a structured candidate profile.
3. Open Search and configure web-search or ATS discovery settings.
4. Click "Find jobs" to run discovery through the Next.js API route.
5. Review ranked jobs in the Jobs tab.
6. Add an OpenAI-compatible API key in Drafts, select a job, and generate tailored answers.

APIs Used

• Greenhouse Board API for public Greenhouse job boards.
• Lever Postings API for Lever-hosted roles.
• Ashby Job Board API for Ashby-hosted roles.
• SimplifyJobs GitHub README feeds for new-grad and internship lists.
• SerpAPI or Google Custom Search API for broader web discovery.
• OpenAI-compatible Chat Completions API for tailored application drafts.

Company Sources

Paste JSON like this into the Company Sources box:

[
  { "company": "Notion", "platform": "ashby", "slug": "notion", "enabled": true },
  { "company": "Palantir", "platform": "lever", "slug": "palantir", "enabled": true },
  { "company": "DoorDash", "platform": "greenhouse", "slug": "doordashusa", "enabled": true }
]

The Next app calls these sources from /api/discover when you click "Find jobs".

New-Grad and Internship Feed

The settings page includes SimplifyJobs feeds:

• New grad: SimplifyJobs/New-Grad-Positions
• Internships: SimplifyJobs/Summer2026-Internships

Set "max age days" to control freshness. For example, 3 shows roles marked around 0d-3d old, while 14 gives a wider list.

Web Job Search

Web Job Search is optional and needs either a SerpAPI key or a Google Custom Search API key plus search engine ID. JobPilot generates search queries from enabled Search Profiles, combines them with any extra queries you add, prefers direct career/ATS domains, excludes noisy aggregator sites, and converts search results into the same job format used by Greenhouse, Lever, Ashby, and Simplify.

LLM Layer

The LLM integration is deliberately a drafting layer, not an autopilot. It builds a compact prompt from your local profile and the target job description, asks for strict JSON, and returns editable drafts. This makes the project more than normal CRUD/web UI work: it combines resume parsing, retrieval from public job feeds, deterministic matching, profile-driven web search, and guarded LLM generation.

Safety rules

JobPilot does not submit applications for you. It ranks jobs, opens application links, and drafts text for review. Matching is deterministic and used for ranking/recommendations, while LLM output is kept as editable draft content.

Resume / Interview Framing

JobPilot | AI-powered Job Discovery Platform

Built a Next.js job discovery platform that turns a resume into a structured candidate profile, discovers new-grad/internship roles from ATS APIs, SimplifyJobs feeds, and profile-based web search, then ranks opportunities with deterministic fit rules.

Integrated an optional OpenAI-compatible LLM drafting layer that generates tailored "why company", AI experience, cover-note, match-summary, and keyword suggestions from the candidate profile plus job description, with JSON-constrained outputs and editable review before reuse.

Tech: Next.js, React, JavaScript, Next API routes, localStorage, Greenhouse Board API, Lever Postings API, Ashby Job Board API, SimplifyJobs feeds, SerpAPI / Google Custom Search, OpenAI-compatible chat completions, deterministic scoring, resume parsing.

Verify

npm test