Muhammad Umar Amanat·Jul 23AI Voice Assistant: A Vendor-Agnostic Architecture for Small HotelsToday, many platforms allow you to build AI voice assistants quickly, helping you to go live in minimal time. However, the real challenge…
Muhammad Umar Amanat·Jun 25AI Reminder Agent That Costs Less Than a Coffee: LangGraph + AWS + TerraformFrom forgotten tasks to daily momentum. A lightweight AI agent built for consistency, and still cost less than a coffee.
Muhammad Umar Amanat·Jan 30Design Once, Scale Safely: A Production Ready AWS RAG Architecture That Avoids Rework CostsDesigned to save engineering time, reduce security risk, and avoid costly re-architecture later.
Muhammad Umar Amanat·Jan 15Enterprise-Grade Amazon OpenSearch: Private Networking & Secure AccessA secure, private OpenSearch architecture designed for enterprise environments where data exposure can cost millions.
Muhammad Umar Amanat·Jul 13, 2025Quick Deployment: AWS Vector Search with TerraformIntegrating cloud services through a user interface might seem convenient at first, but it quickly becomes inefficient when you need to…
Muhammad Umar Amanat·Jul 6, 2025Building Smarter Search: Vector Search on AWS OpenSearch ServerlessAmazon OpenSearch Service is a fully managed search and analytics engine that supports both traditional keyword search and modern vector…
Muhammad Umar Amanat·Jan 14, 2025Simplify SageMaker Batch Scheduling with AWS Lambda & EventBridgeSageMaker transform Job is a powerful tool for generating inferences on large datasets. However, in real-world scenarios, you often need to…
Muhammad Umar Amanat·Jan 9, 2025Deploy Qwen2 as AWS Batch Transform with Custom InferenceQwen2 is Alibaba Cloud’s series of Qwen LLMs. It comes in five sizes: 0.5B, 1.5B, 7B, 57B-A14B, and 72B. Qwen2–7B-Instruct can handle up to…
Muhammad Umar Amanat·Sep 7, 2024QLoRA Fine-Tuning of LlaMA-3–8B on AWS SageMakerPreviously I wrote an article on fine-tuning LLM using the PEFT Technique. I got a good response from the audience, you can find it here…
Muhammad Umar Amanat·Jun 18, 2024Exploring uncommon in Python, part-1: Generator ExpressionMost of us use Python in our daily routine as a software engineer, Data Scientists, Machine learning engineers, etc. But still, several…