Who is behind Oren Stack
A platform engineer who builds the layer that runs AI at scale: provisioning, isolation, control, and dispatch, and the full stack around it.
Built by someone who benchmarks first
Oren Stack started with a benchmark. On 175 real-world queries we measured embeddings at 25% accuracy and hybrid retrieval at 33%. No single technique is the answer. The right tool depends on the job.
So we build the range. Deterministic retrieval when exactness matters, semantic search when similarity matters, and the infrastructure to run either at scale: agent orchestration, control-planes, tenant isolation, ERP and CRM systems, PWAs, MCP servers, and local AI infrastructure.
The result is a catalog of tools and services built around one principle: you own your data, your stack, and your costs.
Let us build something solid
Benchmark first, then build. If you own your data, your stack, and your costs, we will get along.