Hi, I’m Srinidhi Ramanujam, an applied AI architect based in Bengaluru.
I build production GenAI systems for regulated enterprises, especially BFSI workflows where data access, audit, security, latency, cost, and trust all affect the design.
I work across discovery, architecture, data integration, evaluation checks, security controls, and rollout. Recent work includes financial analytics agents, text-to-SQL, agentic ingestion, and evaluation harnesses for enterprise AI products.
Before GenAI, I spent years building production ML systems in telecom. That work included video quality and traffic classification systems, and it led to two granted US patents.
After the demo, teams still need clear data models, measurement, security controls, and systems that users can inspect.
Selected public work
- Lessons From a Year of Shipping LLM Analytics Products for Enterprises: What changes when an LLM analytics product has to survive real workflows and production controls.
- Context engineering experiments: A published experiment on long context, retrieval, structured prompts, latency, and answer quality. The companion repo is on GitHub.
- Quant Magic: A public financial analytics project using natural language to SQL over S&P 500 data.
- Patents: US12069122B2 and US11310162B2.
Contact
For work conversations, LinkedIn is the best route.