Back to Tech landscape

Ollama

Backend

Local LLM runtime for dev and edge

Official documentation (opens in a new tab)

Overview

Ollama runs open-weight models on developer machines or private hosts. ConnectSoft teams use it to prototype RAG flows, compare prompts, and keep sensitive data off public APIs before wiring Microsoft.Extensions.AI or Azure inference endpoints.

Factory templates and SaaS platform programs standardize on .NET 10, Clean Architecture, and Azure-native services—so generated solutions inherit consistent security, observability, and upgrade paths instead of one-off frameworks.

When a client estate or integration boundary calls for this technology, we wire it with explicit ownership, contract tests, and telemetry—without diluting the Factory baseline your long-term platform depends on.

Planning a backend on Azure with predictable releases? Our stack choices are documented, template-backed, and ready to extend—talk to us about mapping your domain to a production-grade .NET platform.

Why we use it:

  • Model pull and version pinned per team
  • Not a substitute for production Azure AI routes
  • Resource limits documented for laptops and CI
  • Same guardrails as cloud LLM usage
  • Aligns with ASP.NET Core hosting and OpenAPI contracts in our templates
  • Observability hooks for Application Insights and OpenTelemetry
  • Clear Factory default versus client-specific adoption boundaries

Ready to work with us?

Let's discuss how our team can help achieve your digital goals