📍 Pune, India ↔ Raleigh, NC | 🧠 Data × AI builder | ⚙️ Platform thinker | 🚀 Principal Data Engineer @ Red Hat
14+ years in data and infra. Spent nearly a decade at Red Hat leading large-scale data engineering, AI/ML platforms, and marketing data systems. Now building open-source tools at the intersection of data pipelines and agentic AI — where the interesting problems live.
- 🧠 memorix-sdk — Core SDK: modular API, YAML config, tiered agentic memory
- 🌐 memorix-server — REST API server exposing MemoryAPI for AI agents
- 🔌 memorix-embedders — Embedding plugins: OpenAI, Gemini, HuggingFace
- 🗄️ memorix-vectorstores — Vector store plugins: FAISS, Qdrant, Chroma
- 📊 memorix-meta — Metadata integrations: DuckDB, Postgres
- 🏭 Streamlit Data Product Factory — Generate dbt models + tests using Gemini 2.5 Flash + LangChain
- 🌊 fluvii — Kafka streams client library for marketing data flows
- 🔁 relay — AI-powered dev workflow: seamless task handoffs + context across Jira, Git, and any IDE
- 🕵️ slack-intel — Fully local Slack history analysis — Ollama RAG, zero data leaves your machine
- 🔗 rlinks — Crowd-sourced bookmark manager with short URL redirects, voting, and AI-suggested tags
- 🤖 askET — RAG chatbot over Red Hat Emerging Technologies content — Gemini + LangChain + FAISS, 100% query success rate
| Org | What it is |
|---|---|
| redhat-community-ai-tools | MCP servers, multi-agent orchestration & AI tooling for Red Hat's open source ecosystem |
| rh-marketingops | Marketing data pipelines, Kafka streaming, and data quality tooling at Red Hat scale |
| purefield-demo-team | RAG apps, AI hackathon starters, and OpenShift demo infra |
| redhat-events | Event-driven tooling, DevSpaces configs, and internal automation |
| PyConPune | PyCon Pune community — Python conference org for Pune, India |
| OpsDataCamp | Data engineering & science learning resources and visualization best practices |
- 📈 qlikreader — CLI/SDK to extract structured data from Qlik dashboards
- 🔗 python-outreach — Python client for Outreach CRM API
- 🪟 dwm — Red Hat's business logic for maintaining marketing data quality
→ Building Memorix AI — agentic memory SDK for AI systems with recall, context, and privacy
→ Patent drafting — around agentic memory systems and AI platform design
→ Open-source — tools that bridge data platforms and agentic AI workflows
→ Writing — sharing learnings on LLM inference, memory, and agent orchestration
"Build tools that solve real problems. Ship them publicly. Let the work speak."
