AI Lab
Agentic AI
Research & Experiments
A playground for ideas, experiments, and prototypes. Here I explore new AI technologies, build proof-of-concepts, and share what I learn along the way. Real experiments. Real code. Real learnings.
- Experiments
- 21+
- Open Source
- 21+
- Hands-on
- 100%
- Learning
- Always
Featured Experiments
Selected projects and prototypes that showcase my ongoing work in AI and agentic systems.
- AI AgentsFeatured
AegiSense AI Assistant
An AI-powered assistant for identity and security operations. Uses RAG, tool calling, and guardrails to answer security questions and automate workflows.
- LangGraph
- FastAPI
- PostgreSQL
- Docker
- Multi-AgentFeatured
Research Crew
A multi-agent system that conducts in-depth research on any topic using specialized agents (researcher, critic, summarizer).
- LangChain
- CrewAI
- OpenAI
- Python
- RAGFeatured
Document Intelligence Lab
Experiments with different RAG pipelines, chunking strategies, and vector databases to find the best setup for real-world use cases.
- LlamaIndex
- Qdrant
- OpenAI
- PDF/Docs
- MCPFeatured
MCP Integrations
Exploring Model Context Protocol (MCP) with real-world tools like Google Drive, Notion, GitHub, and custom APIs.
- MCP
- Python
- Notion API
- GitHub API
- GuardrailsFeatured
AI Guardrails Playground
Testing input/output guardrails, content filters, and policy-based controls for safer AI agents.
- Guardrails AI
- OpenAI
- Policy Engine
- Eval
- EvaluationsFeatured
LLM Evaluation Toolkit
A set of tools and notebooks for evaluating LLM responses using various metrics (relevance, safety, factuality, bias).
- LangSmith
- Ragas
- DeepEval
- Jupyter
All Experiments
Every experiment, prototype and tool in the lab — from quick spikes to production-ready building blocks.
21 experiments
An AI-powered assistant for identity and security operations. Uses RAG, tool calling, and guardrails to answer security questions and automate workflows.
- LangGraph
- FastAPI
- PostgreSQL
Exploring Model Context Protocol (MCP) with real-world tools like Google Drive, Notion, GitHub, and custom APIs.
- MCP
- Python
- Notion API
A multi-agent system that conducts in-depth research on any topic using specialized agents (researcher, critic, summarizer).
- LangChain
- CrewAI
- OpenAI
An agent that can search, read, summarize, and generate insights from multiple sources (papers, docs, web).
- LangChain
- OpenAI
- SerpAPI
Experiments with different RAG pipelines, chunking strategies, and vector databases to find the best setup for real-world use cases.
- LlamaIndex
- Qdrant
- OpenAI
A local web UI for stepping through agent runs — prompts, tool calls, tokens and timings.
- TypeScript
- React
- OpenTelemetry
Testing input/output guardrails, content filters, and policy-based controls for safer AI agents.
- Guardrails AI
- OpenAI
- Policy Engine
A set of tools and notebooks for evaluating LLM responses using various metrics (relevance, safety, factuality, bias).
- LangSmith
- Ragas
- DeepEval
An experimental agent that can help with deployment, monitoring, and troubleshooting using natural language.
- Python
- Docker
- Kubernetes
Specialised agents that review pull requests for bugs, security issues and style, then merge their feedback.
- LangGraph
- GitHub API
- Claude
An AI agent that answers customer queries using company documentation.
- RAG
- FastAPI
- Next.js
Analyzes data, generates insights, and creates visualizations automatically.
- Python
- Pandas
- OpenAI
Benchmarking dense, sparse and hybrid retrieval with re-ranking on real enterprise documents.
- Python
- PostgreSQL
- Qdrant
Generates blog posts, tutorials, and technical documentation.
- LangChain
- OpenAI
- Notion API
A read-only MCP server that lets assistants explore schemas and run safe SQL queries.
- MCP
- TypeScript
- PostgreSQL
Transcribes, summarizes, and extracts action items from meetings.
- Whisper
- OpenAI
- Notion
A general-purpose assistant for daily tasks, research, and learning.
- LangGraph
- MCP
- Tool Use
A drop-in filter that detects and redacts personal data before prompts reach the model.
- Python
- Presidio
- FastAPI
Analyzes market data and provides investment insights (experimental).
- Python
- yFinance
- OpenAI
An agent that helps me learn new technologies with structured plans.
- OpenAI
- Notion
- RAG
Compare prompts and models side by side from the terminal, with diffs and cost estimates.
- Python
- Typer
- Claude
From Experiments to Real-World Impact
These experiments help me stay at the forefront of AI technology and often influence my client work, open-source contributions, and writing.
- Hands-on experimentation with latest AI technologies
- Real-world use cases and practical implementations
- Open source code and detailed documentation
- Learnings shared through articles and tutorials
“The best way to learn AI is to build, break, and build again.”
Follow the Journey
Check out my GitHub for the latest experiments, or subscribe to get updates on new projects and learnings.