Sign up to access an inspiration library of 30,000+ posts
Sriniketh Muralikrishna's Recent LinkedIn Posts
SM
Sriniketh Muralikrishna
@srinikethmuralikrishna
Staff Data Engineer - Health Catalyst | Ambulatory & Healthcare Data
Posts
Sriniketh Muralikrishna
Tech & AI
2mo
Last year I earned my Databricks Data Engineer Associate certification. But I didn't want to stop there — I wanted to build something production-grade. Something with real data, real decisions, and a real agentic layer on top.
So I built TickerPulse - a real-time stock analytics pipeline on Databricks that monitors 15 U.S. equities and generates autonomous trading signals every day.
Here's how the architecture is defined:
- Bronze → Silver → Gold Medallion architecture with idempotent Delta upserts
- Spark Structured Streaming with Auto Loader ingesting 1-minute OHLCV bars from Polygon.io
- RSI, volatility scoring, and cross-ticker correlation computed via window functions and Python UDFs
- LangChain agent powered by Meta Llama 3.3 70B — calls tools, reasons over enriched data, writes structured signals autonomously
- TickerPulse chat UI built with Gradio and deployed as a Databricks App
This is my capstone project for Zach Wilson's and Eumar Dias de Assis's 2026 Winter Databricks Data Engineering Bootcamp — a program that pushed me from knowing Databricks to actually building with it. Thank you for such an amazing 5 week program!
🔗 GitHub: https://lnkd.in/gjgkrby7
81
Sriniketh Muralikrishna
Tech & AI
2mo
Last year I earned my Databricks Data Engineer Associate certification. But I didn't want to stop there — I wanted to build something production-grade. Something with real data, real decisions, and a real agentic layer on top.
So I built TickerPulse - a real-time stock analytics pipeline on Databricks that monitors 15 U.S. equities and generates autonomous trading signals every day.
Here's how the architecture is defined:
- Bronze → Silver → Gold Medallion architecture with idempotent Delta upserts
- Spark Structured Streaming with Auto Loader ingesting 1-minute OHLCV bars from Polygon.io
- RSI, volatility scoring, and cross-ticker correlation computed via window functions and Python UDFs
- LangChain agent powered by Meta Llama 3.3 70B — calls tools, reasons over enriched data, writes structured signals autonomously
- TickerPulse chat UI built with Gradio and deployed as a Databricks App
This is my capstone project for Zach Wilson's and Eumar Dias de Assis's 2026 Winter Databricks Data Engineering Bootcamp — a program that pushed me from knowing Databricks to actually building with it. Thank you for such an amazing 5 week program!
🔗 GitHub: https://lnkd.in/gjgkrby7
81
Sriniketh Muralikrishna
Tech & AI
2mo
Last year I earned my Databricks Data Engineer Associate certification. But I didn't want to stop there — I wanted to build something production-grade. Something with real data, real decisions, and a real agentic layer on top.
So I built TickerPulse - a real-time stock analytics pipeline on Databricks that monitors 15 U.S. equities and generates autonomous trading signals every day.
Here's how the architecture is defined:
- Bronze → Silver → Gold Medallion architecture with idempotent Delta upserts
- Spark Structured Streaming with Auto Loader ingesting 1-minute OHLCV bars from Polygon.io
- RSI, volatility scoring, and cross-ticker correlation computed via window functions and Python UDFs
- LangChain agent powered by Meta Llama 3.3 70B — calls tools, reasons over enriched data, writes structured signals autonomously
- TickerPulse chat UI built with Gradio and deployed as a Databricks App
This is my capstone project for Zach Wilson's and Eumar Dias de Assis's 2026 Winter Databricks Data Engineering Bootcamp — a program that pushed me from knowing Databricks to actually building with it. Thank you for such an amazing 5 week program!
🔗 GitHub: https://lnkd.in/gjgkrby7
81