Preserve provenance at the record level
Reported, submitted, derived, and official information stay distinct, with source URLs and capture context retained for review.
b4joinacompany
website
An evidence-labeled company research product that helps job seekers prepare better questions without turning anonymous reports into verdicts.
- Role
- Product design, evidence modeling, and full-stack engineering
- Scope
- Decision support, cited AI, private research, and versioned data
- Status
- Public research experience live
Problem
Job seekers in Bangladesh often assemble company research from workplace stories, salary submissions, official pages, and interview conversations. Those sources have different levels of reliability, but conventional summaries can flatten them into an unfair good-or-bad judgment.
Constraints
- C1
Personal workplace reports provide context, not verified statements of current company policy.
- C2
Company names and salary records can be ambiguous, so fuzzy matching cannot silently publish a result.
- C3
AI answers must remain bounded by retrieved evidence and still work when no model provider is configured.
Architecture
Versioned source files feed a deterministic research layer. The web product and versioned API consume the same evidence model, while private checkpoints live separately in MongoDB.
01
Source releases
Versioned company, story, comment, and salary evidence records
02
Research layer
Deterministic identity matching, evidence labels, and question ranking
03
Product surfaces
Next.js research briefs, compare, Ask, and a versioned REST API
04
Private workspace
Better Auth and MongoDB checkpoints pinned to evidence revisions
Decisions
Expose unknowns instead of inferring
Missing salary, work-arrangement, or hiring evidence appears as a gap rather than a negative score or guessed policy.
Keep AI optional and cited
Ask retrieves a bounded company-specific evidence set, maps claims to source labels, and falls back to deterministic retrieval when providers are unavailable.
Contribution
- Defined the product methodology and the separation between reports, derived signals, and official facts.
- Built company research, comparison, cited Ask, correction, and private checkpoint workflows.
- Designed the versioned API shared by the website and companion browser extension.
- Implemented optimistic checkpoint revisions and separated private workspace data from the public evidence release.
Outcomes
- Visitors can research a company without creating an account and open original evidence for important claims.
- Prepared questions are generated by fixed rules, while Ask provides cited synthesis or an explicit evidence gap.
- Saved checkpoints retain the exact evidence version and append revisions on successful updates.
- The product supports comparison without producing a synthetic company score or declaring a winner.
Product screenshot
Verified stack
- Next.js
- React
- TypeScript
- Tailwind CSS
- MongoDB
- Gemini API
- Groq API
- Better Auth
- REST APIs
- Zod
- Vitest
- System Design
Working through something similar?
Share your challenge, scope, and timeline to start a focused conversation.