Not because HR-tech looked exciting. Mostly because every company we talked to had the same setup:
- grades in Excel
- 360 reviews in Google Forms
- competency matrices buried in Confluence
- “performance systems” that nobody wanted to open twice
And eventually: “Let’s just build our own thing.”
So we did.
Then this year we made a bigger decision: we rewrote the platform from the ground up — and open sourced it.
Meet HRPulsar
An open source talent management platform for teams where every employee already works with AI tools.
Most HR systems still model companies like it’s 2015: employee → position → annual review.
But that model is breaking. Roles change every quarter. Skills decay faster than job titles. Half the team uses AI agents nobody tracks. And “career frameworks” are often PDFs with better branding.
We built HRPulsar around a different assumption:
Skills are the stable unit. Not positions.
So the platform is competency-first at the core: competency graphs, grade systems tied to behavioural anchors, 360 assessments, internal talent marketplace, development plans, AI fluency tracking, AI workforce registry
Yes, AI workforce registry.
Because “Shadow AI” is becoming a real operational problem: teams use 15 different AI tools, security doesn’t know about half of them, and HR systems pretend none of this exists.
So we added:
- AI tool registry
- employee ↔ AI tool assignments
- oversight levels
- workflow ownership
- audit trails
- workforce maps for hybrid teams (human + AI)
Useful for compliance & security. But mostly because companies should probably know which AI systems are already part of their workforce.
Technically, HRPulsar is
- FastAPI
- Next.js
- PostgreSQL
- Redis
- 200+ API endpoints
- ~1000 automated tests
- Docker-first deployment
- AGPLv3
No “book a demo”. No sales call before seeing the product.
And one thing we decided early: HRPulsar will stay free for individuals and small teams.
Every workspace gets a monthly pool of renewable credits. Enough to run the core platform without turning “try the product” into a budgeting discussion.
And yes — self-hosting is fully supported. Because HR data is sensitive.
And “trust us with your entire workforce structure” is a pretty big ask from a black-box SaaS vendor.
We’re preparing the public GitHub release right now.
*The repository goes live at the end of May. If you want early access, release updates, or just want to watch the project evolve — join the waitlist.
LLMs are useful ONLY for competency matching and draft recommendations
They are not qualified to decide someone’s promotion. We’re very explicit about that in the product.
LLMs are useful for competency matching and draft recommendations.
They are not qualified to decide someone’s promotion. We’re very explicit about that in the product.
Current limitations?
We don’t cover recruiting yet. ATS/recruiting is on the roadmap, but not today.
Would genuinely love feedback from developers building internal tools, people running self-hosted infrastructure, HR engineers, teams experimenting with AI-heavy workflows... and especially from anyone who has ever tried to manage competencies in Excel. We know you exist.
Top comments (0)