The problem
- AI infra is hard: GPUs, K8s, security
- Cloud lock-in, rising costs
- Dev -> prod drift, fragile ops
- Months to reliable launch
The solution
- One-click, self-hosted AI platform
- “Batteries” you turn on as needed
- Identical local -> cloud environments
- Secure by default, no YAML
Customer
- 50-5k employees
- Invest in dev; not tech giants
- Low ROI from AI/DevOps spend
- Frustrated: price swings, complexity, slowdown
- Have favorite cloud/hardware vendor
Market
- AI infra spend surging
- DevOps platforms consolidating
- Self-hosted AI is the gap
- TAM large and growing.
Why now
- AI demand outpaces teams
- Privacy + sovereignty push self-hosting
- OSS is best-in-class
- GPUs now everywhere
What ships today
- Up to date Ollama model inference
- Jupyter Notebooks and Postgres + PGVector
- SSO (Keycloak), TLS, ingress
- Magical MacOSX networking
- Observability: Grafana + metrics
-
Example projects (e.g. Web, Serverless, Local LLM Chat, etc.)
- Runs on Kind locally; scales to cloud.
How it works
- “Batteries” = modular, pre-tuned stacks
- Automation wires deps + security
- GPU auto-detect (NVIDIA CTK, NFD)
- Kubernetes base
Traction
- v1 released (Custom Open/Fair/Available License)
- Hundreds of Local Installs
- Local Installs Converting to Registration/Cloud
- All with no marketing spend
Ask
- Raising $2m to reach $2m ARR in 24 months
- Valuation cap: $20m
- Use: product, GTM, support
- Intros: design partners
Use of funds
- Sales growth
- Support & success
- Product expansion in AI/Marketplace
- Runway: 24 months
Targets
- 7 paying teams in 12 months
- $2m ARR target; NRR > 100% in 24 months
- CAC payback < 6 months
- Time-to-first model < 10 minutes
Why we win
- Open source, no lock-in
- Ease of use means ease of land & expand
- Bring your own cloud/hardware
Business model
- Usage based billing
- Support + SLAs
- Marketplace for add-ons
Go-to-market
- Dev-first, self-serve
- OSS + content (docs, posts)
- Land → expand in teams
- Partners & cloud marketplaces
24-month plan
- Sales playbook
- Hardened enterprise features
- GPU/AI autoscaling across clouds
- Security/compliance packs
Team
- CEO: Facebook/Microsoft/Cloudera Alumni
- Team: Built at scale; shipped OSS
- To Join: Strong business/sales/marketing