TRADITIONAL JOB BOARDS ARE TRANSACTIONAL. FINDER IS ADVISORY.
Most job sites treat you as an application counter to maximize recruiter views. Finder acts as your personal technical career strategist.
Keyword Roulette
Rigid keyword filters reject you if your CV says 'Postgres' instead of 'PostgreSQL', completely ignoring transferable technical depth.
Structured semantic extraction understands core vs. supporting skills, actual years of experience, and role seniority.
Black Box Applications
You submit 50+ applications with zero feedback on whether you were a 40% match or an 85% match, or what skill was actually missing.
Transparent match scoring delivers an explicit fit percentage, detailed alignment rationale, and the exact missing skill gaps.
Static Text Walls
Job descriptions are static text walls that leave you guessing how to position your projects or prepare for the interview.
A real-time AI Career Copilot simulates interview questions, explains requirements, and helps address your specific gaps.
A RESILIENT, FOUR-STEP CAREER INTELLIGENCE PIPELINE
Every step is traceable, deterministic where possible, and protected by resilient fallbacks to ensure you never encounter broken states.
Upload Your Resume (PDF)
Drag and drop your standard text-based PDF. The system performs cryptographic magic-bytes verification (%PDF-) and extracts raw text in memory without third-party data broker storage.
AI Candidate Intelligence
Groq-accelerated models (qwen/qwen3.8-27b) decompose your resume into normalized skills, Core vs Supporting capabilities, seniority level, and verified strengths.
Two-Stage Matching Engine
Deterministic SQL array overlap pre-filters active tech roles with weighted pre-ranking (Core 2.0x, Supporting 1.2x), followed by deterministic re-scoring so rankings stay grounded.
Interactive Career Copilot
Engage with a real-time conversational advisor streaming via Server-Sent Events (SSE). Ask how to address skill gaps, simulate technical interview questions, or optimize your application positioning.
ENGINEERED FOR TECHNICAL PRECISION AND TRANSPARENCY
Every feature in Finder addresses a concrete failure point of traditional recruiting software.
Structured Profile Extraction
Extracts text directly from PDFs via pdf-parse, verifies magic bytes (%PDF-), and validates against strict Zod schemas to distinguish Core (2.0x) vs Supporting (1.2x) skills.
User BenefitInstant, highly structured breakdown of your technical market positioning with zero manual data entry.
Two-Stage Matching Engine
Deterministic SQL array overlap pre-filters candidate pools, weighted pre-ranking selects top matches, and deterministic re-scoring locks the ranking.
User BenefitHigh-accuracy job recommendations with transparent fit percentages, detailed reasoning, and exact missing skills.
Conversational Career Copilot
Streams tokens in real-time via Server-Sent Events (SSE), pre-grounded with candidate strengths and active target role requirements.
User BenefitPractice mock interview questions, clarify ambiguous requirements, and develop strategies to compensate for missing skills.
Dual Auth: Web2 + Web3 (SIWS)
Supports standard email/password SSR session cookies alongside cryptographic Sign-In with Sui (SIWS) personal message verification.
User BenefitConnect with a Web3 browser wallet or traditional email with unified identity continuity and zero lock-in.
Automated External Job Ingestion
Scheduled cron synchronization from Remotive API normalizes remote tech roles with timing-safe constant-time secret authentication.
User BenefitDirect access to active, fresh software engineering and tech opportunities ready for immediate evaluation.
Accessible Discovery Drawer
WAI-ARIA accessible slide-over sheet featuring focus trapping, keyboard navigation, and responsive filter controls.
User BenefitEvaluate job specs, match insights, and apply externally without ever losing your place in the workspace.
EXPLORE THE FINDER WORKSPACE EXPERIENCE
See the actual data structures and analytical output produced by Finder when evaluating technical candidates.
Alex Rivers
Senior Full-Stack Engineer • 6 Years of Experience
Professional Summary
Senior software engineer with 6+ years specializing in Next.js App Router, strict TypeScript architectures, and high-concurrency PostgreSQL data modeling. Proven track record leading front-end modernization and building resilient cloud API services.
Exemplary full-stack architecture depth, proven ownership of mission-critical systems, and exceptional grasp of modern React Server Components and data boundaries.
BUILT ON PROVEN OPEN-SOURCE TECHNOLOGIES
No proprietary black boxes. Finder is transparent about how your data is processed, stored, and protected.
Frontend & Runtime
Next.js 16 • React 19 • Tailwind CSS 4
Modern App Router architecture utilizing Server and Client Components, Base UI accessible primitives, and zero heavy render-blocking animation libraries.
AI & LLM Acceleration
Groq Cloud API • qwen/qwen3.8-27b
Ultra-low latency LPU inference using Groq SDK with automatic secondary fallback (openai/gpt-oss-20b) and deterministic mathematical degradation formulas.
Persistence & Security
Supabase PostgreSQL 15+ • 100% RLS
14 relational tables with strict foreign key constraints, automated auth triggers, and private user-isolated storage buckets for resume PDF documents.
Web3 Cryptographic & Memory Layer
Sui Blockchain • SIWS • Walrus Mainnet
Cryptographic Sign-In with Sui (SIWS) personal message verification, in-memory single-use challenge nonces, and Walrus Memory decentralized encrypted storage via Seal.
VERIFIABLE TRANSPARENCY OVER MARKETING CLAIMS
We do not invent fake testimonials or fabricated user metrics. Finder is built in public under the MIT License so you can inspect every line of code, run the test suites locally, and audit how your data is handled.
476+ Automated Tests Passing
Hermetic unit tests and mocked integration test suites validate matching formulas, Zod validation, Groq resilience, and Walrus memory.
Zero Data Brokerage
Your resume PDF is stored exclusively in private, user-isolated Supabase buckets protected by strict PostgreSQL Row Level Security.
Version-Controlled Prompts
No opaque prompt engineering. Finder's extraction schemas and system prompts are version-tracked and open source in the repository.
Public Specifications
Full documentation for API endpoints, database schemas, matching formulas, and quality gates available for direct inspection.
READY TO GET
YOUR CAREER UNSTUCK?
Upload your resume to extract your structured technical profile, discover verified job matches, and prepare for interviews with an AI Career Copilot.
Web3 Native Authentication
Sign in with Sui (SIWS) or email with zero lock-in