Doorlook
Analyze employers with AI using public signals and employee perspectives.
Implementation notes
Doorlook is an AI assistant that helps users look up employers and get employee perspectives on companies. It uses large language models to summarize public information and provide a general view of what employees think about an organization.
What is Doorlook?
Doorlook is a wrapper around modern AI models (initially GPT‑4o) that asks structured questions about a company and produces a concise overview.
Future Development Plans
Platforms and Services
The plan is to provide insight from a variety of sources, when feasible:
- Newspaper articles
- Glassdoor
- Indeed
- Comparably
- LinkedIn posts
- Public blogs
- Union or worker council pages
AI Models
Future support may include additional models such as Ollama 3.1 and Claude 3.5 for broader coverage and better availability.
Links
GitHub Repository
View the source code on GitHub
github.com/Gogeloo/doorlook
Project snapshot
Problem
People evaluating employers often have fragmented data and no concise way to compare sentiment and risk signals.
What it does
- Collects public context and summarizes employer views with AI.
- Presents a concise, structured employer snapshot.
- Supports exploratory comparisons across organizations.
Use cases
- Candidate pre-interview company research.
- Market/brand perception exploration.
- Rapid briefing on public sentiment indicators.
Stack and architecture
- Next.js web app with LLM-assisted summarization.
- Prompt-driven extraction from public sources.
- Simple repository-based deployment workflow.
Status and roadmap
Current status: Archived. Published 7/7/2024 and last updated 7/7/2024.
- Project archived; no active roadmap.
- Potential revisit if source integrations improve.
- Preserved as a reference implementation.
Links
Related research posts
Browse the security research hub for related writing.