oberservability

Desktop GUI application to explore BullMQ queues stored in Redis.

Prompt and response tracing for LLM workflows

Autonomous log-investigation engine retrieves evidence with hybrid BM25 + vector search (RRF + cross-encoder reranking) and drives an agent that reads, hypothesizes, and probes iteratively to surface root causes.

Reliability-first observability for LLM workflows — trace steps, model calls, retries, and costs

🌊 hexawyn is an AI‑powered Kubernetes assistant that predicts failures, diagnoses issues, and speaks to both SREs and management. Built with hexagonal architecture, OpenTelemetry, and a deterministic semantic layer – 80% infrastructure, 20% LLM.

Production-grade AI underwriting platform demonstrating agentic workflows, hybrid RAG, human-in-the-loop review, evaluation gates, observability, and multi-cloud deployment patterns.

An AI knowledge platform that delivers accurate, context-aware responses through adaptive retrieval, intelligent routing, and enterprise-grade security.

API Traffic Replay & Failure Analysis Platform built with FastAPI, React, PostgreSQL & Docker. Capture, replay, compare and analyze API requests.

Kubernetes monitoring stack using Flask, Prometheus, Grafana, Docker, and Kubernetes for real time observability and metrics visualization.

Internal Developer Platform (IDP) for Modern Cloud Engineering

Agentic AI travel planner powered by Groq's Llama 3.3 70B + LangChain + real-time Tavily search generates personalised day-by-day itineraries via a Streamlit UI.

a development time machine for AI assistants

AI-powered incident investigation platform with multi-agent forensics, evidence-backed RCA, safe remediation planning, and human approval gates.

Complete MLOps pipeline: 6-stage ML workflow,, Kubernetes deployment, Prometheus monitoring, Airflow orchestration, CI/CD automation, Obervability Stacks (Prometheus, EFK)