Shivam Taneja
Consultant,
RBC

ABOUT THE SPEAKER:

Shivam is a Senior Platform Engineer at RBC AI Ops, specializing in high-throughput data access and DevOps. Previously, he led cloud data-pipeline and WebReports initiatives for RBC Investor Services. Outside of work, He enjoys street photography and badminton.

TALK TITLE:

From Data Pipelines to Agent Pathways: Making Data Ready for Agentic AI

TRACK:

Technical/ Engineering Talk

SUB TOPIC:

TBA

ABSTRACT:

Your agents won’t get smarter by ingesting more data, they’ll just burn more tokens.
Enterprises rushing to deploy agentic AI are making a costly mistake: treating LLMs as a universal data processing layer. The result? Exploding token costs, unreliable reasoning, and agents that confidently act on data they don’t actually understand. Intelligent agents need more than access to data; they need data that comes with context, not just content. That means trusted quality signals, lineage that answers “where did this come from?”, semantic context that answers “what does this mean operationally?”, and discoverability that answers “is this even the right dataset?” This talk draws on real-world examples from complex enterprise operations to show how evolving your
data platform — through metadata, governance, and semantic layers — is the difference between agents that reason and act safely, and agents that hallucinate at scale.

WHAT YOU’LL LEARN:

TBA