AI Workplace-Procedures Platform
A LangGraph multi-agent platform that guides HR practitioners through high-stakes workplace procedures, with a human in the loop at every consequential step.
Our role: AI backend + web app (v1 → V2 platform)
The challenge
A confidential Australian client needed to turn high-stakes HR work (investigations, offboarding and employment-relations advice) into a guided software product practitioners could trust, without taking human judgement out of consequential decisions.
What we did
The engagement started small: a single AI-assisted investigation workflow. After we delivered that, the client trusted us to expand it into a full multi-agent HR platform: offboarding, a continuous employment-relations advisor, case management and subscriptions. We built a LangGraph multi-agent backend that walks practitioners through a 12-step workplace-investigation procedure and a 6-step offboarding workflow, with human approval at every consequential step, streaming chat, and AI document generation backed by an evidence matrix. A React case workspace gave CHROs, consultants and business owners their own dashboards behind MFA, with Stripe subscriptions and a retrieval layer over the client's own policy documents.
The result
The result is a system where a small HR consultancy or a large employer can run the same high-stakes procedures (investigations, offboarding, employment-relations advice) with an AI guiding each step and a human approving the consequential ones, so decisions stay consistent and defensible. What began as one agent became the whole platform because the client kept trusting us with more of it. It's in final development; the client is withheld under agreement, which is why some work here is described but not named.
Built with
- Python
- FastAPI
- LangGraph
- MongoDB
- React
- Stripe
In final development; client name withheld under agreement.
Have a project like this?
Tell us what you're building. You'll get a written scope and a fixed quote before any commitment.