Bleenq

Boutique AI/ML consulting + incubator

We build production-grade AI systems — and run our own open-source platform.

Bleenq is a boutique AI/ML consulting firm and incubator. The chat on this page isn't a demo of someone else's tech — it runs on the same platform we open-sourced. Want to see what we build? You're using it.

Talk to our AIRead our writing

Consulting

Taking AI systems from prototype to production — where governance, security, observability, and cost control stop being optional.

How we help →

Platform

An open-source, self-hosted AI platform: one LLM gateway, scoped keys, full tracing, and the boring layers done right.

Incubator

We build our own products on the foundation — knowledge-base bots, a voice assistant, and more. Proof of work, in public.

Latest writing

All articles →

A Control Nobody Tests Is a Belief

A governance control you never test is indistinguishable from a control that doesn't work. This closes the series: guardrails as versioned code, an evaluation suite that runs like unit tests and blocks the merge on a regression, the scheduled jobs that check for inventory drift and shadow credentials and fail loudly, and the honest scorecard that marks its own unfinished items instead of claiming completion. Governance as a living system, not a document.

  • AI Governance
  • Guardrails as Code
  • CI/CD
Read article →

The Model Proposes, the Workflow Disposes

The dangerous moment is when an AI takes an action. This is the deterministic layer that decides whether it may: the model emits an intent, never an action, and a policy enforcement point validates the schema, binds every parameter to the authenticated session, checks a per-tool authorization matrix, and hands execution to a workflow engine that holds the credentials the agent never sees. Including the prompt-injection test that passes even when the model is fully compromised — because 'the model resisted' is not a pass criterion.

  • AI Governance
  • Prompt Injection
  • AI Security
Read article →

You Can't Govern What You Can't See

Enforcement starts with an inventory, because you cannot govern a system you don't know exists. This is how we build the AI inventory and the AI bill-of-materials, why the embedding model and the knowledge corpus belong in that bill as attack surface, how T0–T3 risk tiers make control proportional, and the five signals we use to hunt the shadow AI already running inside the walls — plus why the paved road has to be faster than the goat path.

  • AI Governance
  • AI-BOM
  • AI Inventory
Read article →

Open-source & projects

All projects →

ai-homelab

active

A production-grade, self-hosted AI platform — 40+ services, security and observability from day one.

  • AI Platform
  • Self-Hosted
  • Infrastructure
  • LLM Gateway
MIT
Repo →

discord-curator

active

A Discord bot that surfaces our knowledge bases where the team already talks.

  • Discord
  • RAG
  • Knowledge Base
  • Bot
MIT
Repo — coming soon

Apollo

incubating

A personal AI voice assistant — model, knowledge bases, memory, and tools baked in.

  • Voice Assistant
  • LLM
  • Agents
  • Personal AI
Repo — coming soon