pgflow for AI Workflows
Figure out if pgflow fits your LLM chains, RAG pipelines, or AI data processing. Get architectural guidance for deployment.
Chat with the author
18+ years building production systems. Currently focused on pgflow and AI workflows in Supabase.
Let's talk if you're working on something interesting.
Let's talk if you're working on something interesting.
Hi, I’m Wojciech Majewski
Section titled “Hi, I’m Wojciech Majewski”I love solving challenging problems with code - that’s what drew me to workflow orchestration. pgflow keeps everything in Postgres: no external orchestrators, no extra infrastructure. Just workflows you can actually debug and reason about.
If you’re working with AI/LLM pipelines in Supabase and want database reliability with TypeScript ergonomics, let’s talk.
Why pgflow
Section titled “Why pgflow”⚡ <100ms job startup - Near-instant execution
🔌 Run from anywhere - SQL (triggers, pg_cron), browser, or RPC
📡 Realtime progress tracking - Live updates via websocket
👁️ Full observability - Complete history in SQL tables
✨ No boilerplate - Skip the cron/queues/Functions plumbing
📦 Zero new infrastructure - Everything in Supabase
What I can help with
Section titled “What I can help with”Supabase + AI Architecture
Full-stack Supabase consulting for AI applications and database-first architectures.
Proof of Concept
Build a working pgflow prototype for your specific use case.
pgflow is a good fit if you
Section titled “pgflow is a good fit if you”- Already use Supabase for your application
- Have AI/LLM workloads that need orchestration
- Want everything in PostgreSQL (no external services)
- Need to see exactly where and why workflows fail
- Understand this is early but powerful software
Let’s talk
Section titled “Let’s talk”15-minute technical discussion to see if pgflow fits your needs.