Insights, tutorials, and updates from the Akruti Consulting team on web development, AI, and modern tech.
Discover how GitHub Copilot has evolved from a code completion tool to a powerful platform for building and orchestrating autonomous AI agents that can plan, execute, and learn from your development workflow.
February 12, 2026
9 min read
AI agents are everywhere in the headlines, but what do they actually do? This primer cuts through the noise and explains agents through a developer's task list that runs itself — then maps the 12-day series ahead.
May 11, 2026
5 min read
Stop reading about agents and start building one. This post wires up a Claude agent that calls a real weather API, and walks through every message, tool call, and loop iteration so you can see exactly how the model drives itself.
May 12, 2026
9 min read
The OpenAI Agents SDK (released March 2025) flips the mental model from 'chat loop' to 'agent object'. In this post we build a two-agent pipeline with tools, a handoff, and a guardrail — in about 60 lines of Python.
May 13, 2026
5 min read
Both Anthropic and OpenAI support tool/function calling, but the APIs differ in meaningful ways. This post builds the same agent with both SDKs and surfaces the key differences in schema format, invocation style, and error handling.
May 14, 2026
5 min read
Agents without memory are amnesiac — every conversation starts from zero. This post explains the three memory tiers (short-term, long-term, episodic), when to use each, and includes a working key-value memory store you can wire into any Claude or OpenAI agent today.
May 15, 2026
5 min read
Every AI framework invented its own way to expose tools to models — until MCP. Anthropic's Model Context Protocol is rapidly becoming the USB-C of agent tool connectivity. This post builds a minimal MCP server in TypeScript and shows you how to connect it to Claude Desktop or GitHub Copilot.
May 17, 2026
5 min read
The OpenAI Assistants API is a fully-managed agent runtime: it persists conversation threads, handles context-window overflow, and bundles hosted tools like Code Interpreter and file search. This post builds a working data-analysis assistant that reads a CSV and answers questions about it.
May 18, 2026
5 min read
Retrieval-Augmented Generation (RAG) turns a static document collection into a dynamic tool any agent can query. This post builds a full local RAG pipeline — chunking, embedding, storing in ChromaDB — and wires it into a Claude agent as a callable tool, all without any hosted infrastructure.
May 19, 2026
5 min read
A single agent trying to research, write, and review at the same time produces mediocre results in all three. LangGraph lets you split those concerns across specialised agents and wire them together as a directed graph. This post builds a researcher–writer pipeline that produces a publishable blog draft.
May 20, 2026
5 min read
A prototype that works on your laptop is not a production service. This post covers the plumbing decisions that separate the two: structured output with Zod and Pydantic, retry and fallback strategies, trace logging, and how to choose between serverless and long-running deployment models.
May 21, 2026
5 min read
Traditional unit tests break down for agents — the output is non-deterministic and the logic is distributed across LLM calls and tool invocations. This final post in the series covers four practical evaluation strategies: unit-testing tools in isolation, trace-based evaluation, LLM-as-judge scoring, and golden-dataset regression testing.
May 22, 2026
6 min read