> writing
All Posts
9 posts
This Week in AI — 26 April 2026 | Weekly AI & ML Roundup
Top AI and machine learning developments for the week of 19 April 2026: 29 stories across LLMs, AI agents, research, and industry news curated by Jag Patel.
Local-First AI Security Auditor: Evidence, Not Excuses
A local-first AI-powered security auditor that ingests SARIF/JSON from any tool, deduplicates findings, and outputs a single, evidence-based Markdown report—no hallucinations, no data leakage.
Building a Local-First AI Knowledge Base — Chat with Your Files, Offline
I built a unified local AI chat interface that lets you query PDFs, Word files, PPTs, markdown, and notes using RAG — fully on-device with Gemma 4 (2B) via Ollama, ChromaDB, and NetworkX. No cloud. No data leaves your machine.
🚀 Building a Production-Grade Python Logging SDK for Distributed Observability
A lightweight, plug-and-play Python Logging SDK that standardizes structured observability across AI/ML pipelines, distributed systems, and backend services — with automatic context injection, sensitive data redaction, and OpenTelemetry support.
Building a Local MCP Server to Automate Daily DevOps Tasks
I built a fully local MCP (Model Context Protocol) server using FastAPI and Phi-3 Mini to automate repetitive cloud and DevOps tasks — no cloud, no API keys, entirely on-device.
Building an AI Decision Engine with Mistral, FastAPI & PromptFlow
I built a dynamic decision engine that evaluates user data against JSON-defined rules and provides AI-driven reasoning using Mistral-7B-OpenOrca via GPT4All — with PromptFlow for workflow orchestration, human-readable PDF reports, and a full CI/CD pipeline on Azure App Services.
Building a Local AI Billing Chat with FastAPI and Ollama
I built a local AI-powered billing chat that reads cloud cost CSVs (AWS, Azure) and lets you query cost changes, new services, and team breakdowns in plain language — no external APIs, fully on-device with Ollama and FastAPI.
Building an AI Decision Engine with Mistral, FastAPI & PromptFlow
I built a dynamic decision engine that evaluates user data against JSON-defined rules and provides AI-driven reasoning using Mistral-7B-OpenOrca via GPT4All — with PromptFlow for workflow orchestration, human-readable PDF reports, and a full CI/CD pipeline on Azure App Services.
🚀 A language-agnostic API & AI/ML E2E testing framework
Building a plug-and-play reliability engine for modern API and AI/ML systems.