Mr. Ganapathi Yeleswarapu

Trustworthy Autonomy: RPA-Driven, Real-Time Data & ML for FinTech

Abstract:

FinTech automation is moving from scripted RPA to agentic, policy-aware systems that blend LLMs, streaming intelligence, and human oversight. This talk maps the frontier: LLM copilots/agents orchestrating end-to-end workflows; event-native automation where transactions and alerts trigger autonomous decisions; and real-time feature fabrics uniting graph + vector signals for fraud, AML, and risk. Governance-as-code is evolving into policy LLMs that interpret controls and generate audit narratives. Reliability hinges on AI observability (live drift/bias/stability) and self-healing pipelines that auto-roll back or re-route. Privacy advances—federated learning, synthetic data, confidential compute, fine-grained consent—enable collaboration without exposing PII. Edge runtimes push on-device inference; cloud continuous evaluation and sandboxed agents gate releases. The result: bots/agents as last-mile executors within guardrails, human supervisors for exceptions, and systems that learn continuously.

Profile:

Ganapathi Yeleswarapu is a seasoned Senior Data Engineer with over 18 years of experience specializing in Agentic AI, Google Cloud Platform (GCP), AWS Glue and ETL development. He has led large-scale data warehouse migrations, notably transitioning legacy systems from Oracle and Informatica to GCP using BigQuery, Dataflow, and Airflow. Ganapathi has deep expertise in real-time data streaming, fraud detection, and compliance reporting, with hands-on experience integrating data from platforms like Kafka, Couchbase, and Salesforce.
His domain knowledge spans financial compliance, including Suspicious Activity Monitoring (SAM) and Enterprise Risk Case Management (ERCM), with a strong track record in optimizing data pipelines for performance and cost efficiency.