Uday BKV

14 pieces, newest first

Archive

Every write-up and note on this site in one place — AI engineering, green software, and the odd hardware tinkering session.

2026

  1. Agentic AI Multi-Agent vs Agent to Agent CommunicationMulti-agent system is the architecture.Agent-to-agent communication is the plumbingNote
  2. Copilot adoption dies in the onboarding gapAccess isn't adoption. The squads that stuck with it were the ones we paired with in week one.Note
  3. Building a Production ready Agentic RAG Pipeline: TF-IDF, HNSW, LSH, CAG, guardrails and MoreThis post outlines a potentially effective approach to user queries by implementing a Agentic Retrieval-Augmented Generation (RAG) strategy,and 10-guardrail safety sys...AI Engineering
  4. Building a Production ready RAG Pipeline: TF-IDF, HNSW, LSH, CAG, guardrails and MoreThis post outlines a potentially effective approach to user queries by implementing a Retrieval-Augmented Generation (RAG) strategy, and 10-guardrail safety system.AI Engineering
  5. Developing powerful Agentic AI through GitHub Copilot’s custom agents.In this post, we’ll explore what Copilot Custom Agents are, the problem they solve, and how you can automate multi-agent workflows using Python.AI Engineering

2025

  1. Evaluating Green IT Metrics for Microservices, utilizing the KeplerThe majority of organizations are deploying and publishing Open APIs and microservices in both public and private cloud environments. Most cloud providers are followin...Green Software
  2. Develop MCPServer with a focus on eco-friendly technology practices.The field of AI Engineering is rapidly evolving and expanding to explore new possibilities. This growth is marked by frequent updates to models, integrations, upgrades...Green Software
  3. Embracing Green AI for a better tomorrowGreen AI: Green AI promotes the creation of sustainable and energy-efficient AI and machine learning models, aligning technology with environmental responsibility.Green Software
  4. AI Model Performance metricsthese metrics are important for model monitoring and improvementAI Engineering

2024

  1. Configuring Wifi in ESP32 WORM using codeRecently, I have been delving into a specific use case that involves consuming a voice REST endpoint using the ESP32 microcontroller. This task requires not only utili...Field Notes
  2. Data mocking using FakerIdeally, test data is of priority and the project teams always face an issue in getting the relevant and realistic test data for pre-production activities. More issues...Field Notes
  3. Unlock the power of offline speech recognition with the VOSK API!Vosk API Overview: A powerful, offline speech recognition framework for converting spoken language into text without needing an internet connection.AI Engineering
  4. Deploy a simple supervised machine learning algorithm in Azure as a function.This post outlines the development of a supervised machine learning model using the DecisionTreeClassifier algorithm. We will cover training the model on a specific da...AI Engineering
  5. What got me pondering about Green Software?Over the years I’ve seen some posters and online flyers and information about climate change, net zero and sustainability and never thought about thinking along these ...Green Software