Honestify
← Back
Knowledge Basev1.3.1

Knowledge Base Launch

Knowledge Base Launch (2026-01-06): The Knowledge Base is live: a private, searchable corpus of your documents, crawled pages, and notes that grounds AI answers in your actual work.

·4 min read·Honestify Engineering

Feature area: Knowledge Base

On this page

This release covers Knowledge Base Launch Version 1.3.1, shipped 2026-01-06. Status: shipped. No breaking changes.

Summary

The Knowledge Base is live: a private, searchable corpus of your documents, crawled pages, and notes that grounds AI answers in your actual work.

Generic AI interview prep hallucinates project details. Without a personal corpus, answers sounded impressive but inaccurate under follow-up. We launched a first-class Knowledge Base with upload, crawl ingestion, chunk indexing, and retrieval wired into chat and profile generation.

What Changed

Document upload

New

Markdown, plain text, and PDF sources added to personal index.

Semantic search

New

Vector search over chunks with metadata filters.

Source management

New

View, disable, or delete indexed sources from dashboard.

Answer accuracy (self-reported)

Before

62%

After

88%

Profiles with 3+ KB sources

  • Document upload — Markdown, plain text, and PDF sources added to personal index.
  • Semantic search — Vector search over chunks with metadata filters.
  • Source management — View, disable, or delete indexed sources from dashboard.

Why We Built It

Generic AI interview prep hallucinates project details. Without a personal corpus, answers sounded impressive but inaccurate under follow-up.

We prioritized this work because user interviews consistently surfaced this as a blocker to daily use. The fix needed to be durable—not a patch—so we addressed root causes in Knowledge Base rather than symptoms alone.

Engineers, recruiters, and hiring managers all benefit when Honestify behaves predictably in production. This release reflects that bar.

User Impact

Answer accuracy ratings (user-reported) improved 41% when KB had 3+ sources.

AudienceHow you benefit
EngineersFaster profile setup, clearer AI answers, less manual rework
RecruitersMore complete profiles and reliable share links when candidates use Honestify
Founders / hiring managersBetter signal on candidate preparation and skills alignment
Platform engineersPerformance headroom for upcoming features

Relevant skills: rag, embeddings, vector databases, system design. Target roles: ai engineer, backend engineer, full stack engineer.

Technical Highlights

  • pgvector for embedding storage
  • Hybrid keyword + vector retrieval
  • Source-level ACLs (private to user)
  • Re-index on document update

Prompt and retrieval changes versioned separately with eval gating before production promotion.

Before

Knowledge Base Launch: before vs after

Before

AI drew only on manually entered profile fields—thin context for senior engineers with deep portfolios.

After

Documents and crawled pages feed retrieval; chat and narratives cite indexed sources.

Users moving from the previous experience should notice documents and crawled pages feed retrieval; chat and narratives cite indexed sources.

Screenshots

Future Improvements

What we are building next

  • Team knowledge bases for hiring managers
  • Version history per source
  • Automatic skill tagging from KB content

Known limitations

  • · 10 MB per document on free tier
  • · Re-index latency up to 2 minutes for large PDFs

Feedback welcome: Reply via in-app feedback or support—especially if you hit edge cases we did not cover in this release.

This update connects to other Honestify work:

Create your own AI profile

Upload your resume, add expertise, and share a profile link beside LinkedIn so recruiters can ask follow-up questions before the interview.

Create your AI profileUpload your resume