AboutBlogsContact

© 2026 Sayan Dey . All rights reserved.

AI LAB

A playground where I explore AI systems , autonomous agents , and intelligent developer tools.

AI AgentsLLM ToolsExperimentsResearch
Explore Experiments ↓View GitHub

What is AI Lab?

AI Lab is my experimental space where I build and test AI-powered systems. I explore large language models, autonomous agents, and developer tools that automate workflows and enhance productivity.

Experiments & Tools

Mini RAG
Mini RAGtool

A modular Retrieval-Augmented Generation (RAG) system built from scratch to explore document indexing, semantic retrieval, prompt engineering, streaming responses, and production-ready AI architectures.

pythonfastapigroq

AI Agents

Mini Agent

A modular AI agent framework built with a TypeScript monorepo and a Python AI runtime. Designed for provider abstraction, streaming, tool calling, and extensibility, making it easy to integrate new models, tools, and agent capabilities.

typescriptpythonfastapiturborepo

AI System Architecture

Most AI experiments in this lab follow a layered architecture combining interfaces, backend orchestration, AI models, and scalable infrastructure.

Interface Layer

User-facing interfaces where AI tools, chat systems, and experiments are interacted with in real-time.

Next.jsTypeScriptTailwind CSSshadcn/ui

Upcoming Experiments

Future explorations planned for the AI Lab. These ideas push the boundaries of intelligent systems and automation.

huggingface
qdrant
langchain
docker
rq
valkey
GitHub
NeuroText MLP
NeuroText MLPresearch

Building a character-level Multi-Layer Perceptron (MLP) language model from scratch in PyTorch. This project extends the Bigram model by learning character embeddings and predicting the next character using multiple previous characters as context.

pythonpytorchjupyter-notebookhugging-face-datasetsmatplotlib
GitHub
NeuroText Bigram
NeuroText Bigramresearch

Exploring language modeling from first principles by implementing both a statistical Bigram model and a neural-network-based Bigram model in PyTorch to learn how machines generate text one character at a time.

pythonpytorchjupyter-notebookhugging-face-datasetsmatplotlib
GitHub
ollama
groq
pnpm-workspace
commanderjs
Coding Agent

An autonomous coding agent built with Python and Groq that solves programming tasks through iterative tool calling. The agent plans one action at a time, observes tool outputs, and continues reasoning until the requested task is completed.

pythongroq-apipydanticdotenvjsonagentic-ai
Weather AI Agent

An autonomous AI weather assistant built with Python and Groq that reasons through user requests, invokes external weather tools when needed, observes the results, and generates accurate natural-language responses using a ReAct-style workflow.

pythongroq-apipydanticjsonrequests

Orchestration Layer

Handles request flow, prompt construction, API calls, and coordination between different AI components.

Next.js API RoutesServer ActionsNode.js

AI Execution Layer

Core intelligence powered by LLMs for reasoning, generation, and tool-based workflows.

OpenAI APIPrompt EngineeringFunction Calling

Memory & Infrastructure

Manages state, caching, and scalability for AI interactions and real-time systems.

Redis (Aiven)WebSocketsVercel