Author: Optimized AI Conference

https://youtube.com/embed/cwBMavL7518
Victor Calderon from Intuit breaks down the journey from prototype AI agents to production-ready systems powering TurboTax and QuickBooks. He covers critical production challenges including multi-tenancy, retry logic, timeout handling, and scalability under peak loads of 300,000 users per hour, offering practical architectural patterns for reliable agentic deployments.

✅ Connect with Victor – https://www.linkedin.com/in/vcalderon/

✅ Connect with Optimized AI Conference on LinkedIn – https://www.linkedin.com/company/oaiconference/

✅ Visit Optimized AI Conference Website – https://www.oaiconference.com/

https://youtube.com/embed/NsUl_hQUot0
Beverly Wright, a data science and AI veteran since 1991, draws on experience at Truist, Southern Company, and Cox Communications to address bridging the gap between enterprise AI initiatives and tangible business outcomes. She covers strategic frameworks for translating AI capabilities into real organizational decisions through her academic and industry leadership perspective.

✅ Connect with Beverly – https://www.linkedin.com/in/drbeverlywright/

✅ Connect with Optimized AI Conference on LinkedIn – https://www.linkedin.com/company/oaiconference/

✅ Visit Optimized AI Conference Website – https://www.oaiconference.com/

https://youtube.com/embed/MKQEQkGha8c
Stanislav Stolpovskiy, search and ML architect at Grid Dynamics, shares hard-won lessons from building enterprise-scale multi-agent systems. He discusses why most enterprise AI projects fail to reach production, outlines proven design principles for agent mesh architectures, and highlights what worked and what didn’t across complex agentic deployments.

✅ Connect with Stanislav – https://www.linkedin.com/in/stanislav-stolpovskiy-260755140/

✅ Connect with Optimized AI Conference on LinkedIn – https://www.linkedin.com/company/oaiconference/

✅ Visit Optimized AI Conference Website – https://www.oaiconference.com/

https://youtube.com/embed/BZbKiDDm7Jw
NVIDIA solutions architects Avinash Ahuja and Vivek Kolasani explore scaling agentic AI infrastructure. They trace the evolution from stateless LLMs to tool-augmented agent systems, then demonstrate how NVIDIA Dynamo and Kubernetes enable disaggregated, scalable inference architectures for production agent workloads across hyperscaler environments.

✅ Connect with Avinash – https://www.linkedin.com/in/avinashahuja/
✅ Connect with Vivek – https://www.linkedin.com/in/vivek-kolasani-b66ba823/

✅ Connect with Optimized AI Conference on LinkedIn – https://www.linkedin.com/company/oaiconference/

✅ Visit Optimized AI Conference Website – https://www.oaiconference.com/

https://youtube.com/embed/HSfLUdCjFSA
AWS developer advocate Darko Mesaroš introduces Strands, an open-source framework for building production-ready AI agents. Through live coding demonstrations, he covers agent architecture fundamentals including tool integration, memory management, and evaluation strategies, all designed to work across any cloud provider or model vendor using Python or TypeScript.

✅ Connect with Darko – https://www.linkedin.com/in/darko-mesaros/

✅ Connect with Optimized AI Conference on LinkedIn – https://www.linkedin.com/company/oaiconference/

✅ Visit Optimized AI Conference Website – https://www.oaiconference.com/