Lessons Learned Building Agentic Systems

Lessons Learned Building Agentic Systems

Friday, November 13, 2026 • 8 to 10 AM PST / 11 AM to 1 PM EST

Can’t attend live? Register anyway and we’ll send you the full session recording.

Reserve Your Spot

All registrants will receive the session recording.

We respect your privacy. Your information will only be used for this webinar and related communications from TechEmpower.

Models keep getting better, but the systems we build on them keep getting harder to make reliable. Applications now carry far more context, hand work between multiple agents, and act with more autonomy against real systems and real customer data. Each of those adds new ways to fail. Many engineering leaders, including the ones presenting, are working through what it actually takes to deliver high quality, reliable agentic applications in production.

In this session, four senior technical leaders from LendingTree, Comcast, ReFocus AI, and Civio share the lessons they learned the hard way building agentic systems that real users depend on.

What You'll Learn:

Practical patterns for assembling the right context for the right task, person, and moment and why context engineering trumps prompt engineering.
How to keep state and conversation context intact across agent handoffs, and why guardrails need that context too (without adding latency).
How to design for failure instead of the happy path: recovery, validation, and treating autonomy as a bounded capability, with deterministic code and humans where correctness matters most.
How to effectively move evaluation and observability into the main loop - along with appropriate trace logging, curated eval datasets, LLM-as-judge scoring, and measuring the system rather than the model.
How to give agents controlled access to real systems and enterprise data: authority, memory, governance, and least privilege.
Why the model is becoming the least differentiated part of the system, and how shared capabilities and infrastructure make the fifth agent faster to ship than the first.

Join fellow CTOs and senior engineering leaders for a pragmatic look at how you can engineer more reliable agentic systems.

Tony Karrer (Moderator)

Tony is the Founder/CEO of TechEmpower, a 20-person software / AI development services company. Tony has worked as a fractional CTO at over 40 different companies including as the original CTO at eHarmony helping them to get to a $110M financing in their 4th year. Tony has worked on AI solutions for more than two decades across a wide variety of industries and is currently the fCTO of AI companies in healthcare, legal and insurance. Tony is also cofounder of the LA CTO Forum, the largest CTO professional organization in the US.

Eric Hanson (Speaker)

Eric Hanson is an AI Platform Architect at LendingTree, where he leads the AI platform architecture used to scale AI assistant workflows and agentic systems across the company. He co-authored the AWS writeup on LendingTree's multi-agent mortgage assistant built on Amazon Bedrock, and his current focus is the layer underneath it: capability registries, MCP tool contracts, shared context services, and evaluation. He also writes The Data Lead, a newsletter on enterprise AI architecture.

Bhavin Parikh (Speaker)

Bhavin Parikh is a Senior Principal AI Architect at Comcast, where he leads the architecture and delivery of enterprise-scale AI platforms and agentic systems for service reliability and operations. He helped design and build Comcast's Reliability Intelligence platform, which uses AI to support incident triage, investigation, resolution, and post-incident analysis, and is a three-time recipient of Comcast's TPX Innovation in Action Award for Generative AI. His current work centers on agentic AI, context engineering, multi-agent orchestration, knowledge graphs, and memory.

Abbas Raza (Speaker)

Abbas Raza is CTO of ReFocus AI, where he leads product engineering and the development of production AI systems that automate complex business workflows. He has more than 20 years of experience building and scaling enterprise software, distributed systems, data platforms, and globally distributed engineering organizations, including leadership roles at SAP, AtlasHXM, and SimplifyVMS. His current work focuses on agentic AI, computer-use agents, AI/ML, reliability, security, and the engineering challenges of putting AI into production.

Anindo Banerjea (Speaker)

Anindo Banerjea is a 3x CTO in Enterprise SaaS & AI/ML with over 20 years in engineering leadership. He introduced AI/ML at MyTime and Archer, led SaaS transformation at MetricStream, co-founded TBD Networks, and is an advisor at AISquare. Anindo holds a Ph.D. in Computer Science from UC Berkeley.