Product Case Study
reqon.appReqon
A self-hosted, AI-assisted job-search CRM — structured pipeline management, a deterministic scout, and optional AI assistance that never takes control from you.
Job searches scatter across spreadsheets, browser tabs, and email. Reqon turns that into a structured product system — spanning a web dashboard, an iOS/iPadOS companion app, and a Chrome extension.
Product
Self-hosted job-search CRM, 3 surfaces
Scale
150+ opportunities tracked
Stack
React Native + Chrome Extension + AI
Outcome
Solo-built, production-shipped
Representative Product Demonstration — This page recreates Reqon's real interface using original components and synthetic data. Company names, roles, statuses, and figures shown are entirely invented for illustration and do not reflect any real job search.
Try It
Your pipeline, one board
Switch tabs to see Today's action items, browse by stage, or check Analytics. Click any role to open its detail panel — everything below is a working, original recreation of the real interface, running on synthetic data.
Open
74
Applied
52
Follow-up due
31
Rejected
18
Interviews
3
Offers
0
Total roles
158
Tier A
41
Response rate
5.8%
Avg days→contact
3.4
Action needed — your daily loop: discover → verify → apply → follow up
Count: 2
New since last run
Added since the last scout run, not yet applied
Count: 5
Needs verification
Unverified links — confirm the live posting first
Count: 19
Apply next
Tier A/B · verified · still open · not applied
Count: 12
Tier A · not applied
Top-tier roles you haven't applied to yet
Count: 31
Follow-up due
Active applications gone quiet past your threshold
Count: 6
Recently closed
Postings detected closed — review and archive
Count: 3
In interviews
Active interview-stage conversations
Case Study
The story behind the product
The Insight
Job searches scatter across spreadsheets, tabs, and email
Good-fit roles get missed, follow-ups get dropped, and application decisions become inconsistent — not because the roles aren't there, but because nothing holds the whole picture together.
The Product Thesis
A structured pipeline, with AI that assists but never decides
Roles are captured, scored, and tracked through a repeatable pipeline. A deterministic scout does the discovery; AI assistance is optional, reviewable, budget-capped, and never auto-submitted — the system helps make better decisions, it doesn't make them for you.
The Outcome
A real, self-hosted system — not a portfolio exercise
Reqon is a production product spanning a web dashboard, an iOS/iPadOS companion app, and a Chrome extension, actively used to track a real job search end to end.
My Role
Sole builder — product, design, and engineering
Product Leadership
- Identified the fragmentation across job boards, spreadsheets, email, and personal notes as one workflow problem, not a tooling gap
- Defined the deterministic-first, AI-optional product philosophy — scoring and recommendations are reviewable, editable, and budget-capped, never auto-submitted
- Designed the pipeline model: lifecycle stages, hygiene lanes, and an expected-value-ranked apply-next queue
- Scoped and sequenced three companion surfaces (web, iOS/iPadOS, Chrome) around distinct jobs — command-center management, quick review, and capture
Technical Direction
- React Native / Expo companion app for iOS/iPadOS
- Chrome extension for posting capture, inline scoring, and autofill
- Deterministic multi-ATS scout polling public board APIs (Greenhouse, Ashby, Lever, and others), no API key required
- Optional AI assistance on the OpenAI Responses API, with per-call token metering and daily budget caps
Companion Surfaces
The same board, wherever you need it
The Chrome extension captures postings and scores them inline while you browse. The mobile companion keeps the same pipeline available for a quick review, synced to the same self-hosted board.
Not on your board
(8) New Product roles for you | LinkedIn
Keyword coverage
27%
158
Roles tracked
74
Open / not applied
Top opportunities to apply
Senior PM, Growth
Northwind Analytics
Principal PM, Platform
Fernbridge Labs
Group PM, Data
Lumen Robotics
Reqon Clip — Chrome extension (recreated, synthetic data)
iOS/iPadOS companion app (real screenshot)
Outcome
What happened
- Actively tracks 150+ real opportunities across a real job search, end to end.
- Shipped three companion surfaces — web, iOS/iPadOS, and Chrome — from a single self-hosted backend.
- Established a deterministic-first, human-controlled AI pattern: every AI output is reviewable, editable, and never auto-submitted.
- Solo-built and production-shipped, including a drafted App Store listing.
What I'd Build Next
- Deeper Gmail-based triage — auto-classifying more recruiter reply types beyond the current rejection/interview detection
- Expanded ATS coverage as more boards adopt public APIs
- Calendar-integrated interview scheduling, tied directly to the pipeline's interview stage
- A public-facing, shareable version of the analytics view for career coaching use cases
Why This Matters
A job search is a product problem — structured data, a repeatable workflow, and just enough automation to remove the busywork without removing your judgment. This is what that looks like when you build the tool you actually need.