Projects & Experiments

A collection of my work exploring Agentic AI, Generative AI, and distributed systems.

Showing 12 of 12 projects

November 2026

A from-scratch agent harness built from scratch in order to deep dive into harness engineering.

AgentsAgent HarnessContext EngineeringSubagentsDockerSandboxingLangSmithEvaluationOpenAIPythonUdemy

May 2026

An autonomous CLI agent that takes a natural language coding request and delivers a complete, reviewed Python project — planning, writing, code-reviewing, and fixing in a single workflow.

deepagentsLangGraphAgentsCode GenerationPython

April 2026

A LangGraph-powered personal finance chatbot with tier-based model routing, structured output, multi-turn memory, and streaming support.

LangGraphLangChainAgentsPython

April 2026

A LangGraph multi-agent travel planner where a Supervisor Agent coordinates specialized Flights, Hotels, Activities, and Itinerary subagents to build complete trip plans from a single request.

LangGraphLangChainMulti-AgentSubAgentsPython

April 2026

Production-ready LLM observability using Langfuse — trace every call, cut costs with smart model routing and semantic caching, and stay compliant with built-in PII redaction.

LangfuseLangchainPythonAnthropicOpenAIDebuggingPrivacyOpentelemetryChromaDBCost Optimization

February 2026

End-to-end RAG system with document chunking, retrieval, and LLM evaluation.

RAGGraphRAGAgenticRAGLangChainEvalsLLM-as-a-JudgeMRRNDCGVector StoreEmbeddingsPython

February 2026

QLoRA fine-tuning of LLaMA 3.2 for product price prediction with W&B tracking.

QLoRALLaMA3.2SFTPeftTransformerPythonQuantizationHyperparameters tuningWeight & BiasesNvidia GPU A100

February 2026

Build an autonomous system of agents to find online deals and notify users.

ModalAgentsPython

January 2026

Understanding LLM reasoning model architecture through Sebastian Raschka's book.

Reasoning ModelsLLMReinforcement LearningDistillationModel EvaluationInference Scaling

September 2025

Step-by-step guide to training and fine-tuning Language Models from scratch.

LLMFine-tuningTrainingRLHFPython

August 2025

Conversational AI assistant for career-related questions using LLMs and RAG.

LLMGenerative AIPythonTool callRAG

June 2025

Multi-agent trading system with 4 autonomous agents powered by MCP servers and 44 tools.

Agentic AIMulti-AgentMCPOpenAI SDK

Interested in collaborating?

Let's discuss AI architectures, team leadership, or potential partnerships.

Get in touch