The VextoraTech Engineering Blog
We write about what we build. AI, full-stack patterns, DevOps, and the occasional hard lesson.
Building a RAG Pipeline from Scratch with ChromaDB and LLaMA 3.2
How we built a local-first RAG system with citations, embeddings, and zero API cost.
Building a RAG Pipeline from Scratch with ChromaDB and LLaMA 3.2
How we built a local-first RAG system with citations, embeddings, and zero API cost.
Why We Use the Repository Pattern in Every FastAPI Project
The architectural pattern that keeps our backends testable, swappable, and sane.
RBAC Done Right: 4 Roles, 16 Permissions, Zero Confusion
A pragmatic role-based access control schema you can ship on Monday.
Local AI vs. API: When to Use Ollama Instead of OpenAI
A cost, latency, and privacy comparison from real client projects.
Docker Compose for Full-Stack Projects: Our Production Template
The compose file we copy into every project, annotated.
Designing for Developers: Building UI That Engineers Actually Use
Lessons from designing dashboards used by engineering teams.
Mermaid.js + AI: Generating Diagrams from Natural Language
How DiagramAI Studio turns one sentence into a system diagram.
JWT Auth in FastAPI: Our Battle-Tested Implementation
Refresh tokens, rotation, and revocation — the production setup.
Get our engineering posts in your inbox.
No fluff, no spam — just stuff we'd want to read.