DataPhilly Tech Talks: Adaptive AI Systems and Decision-Centric Governance

🚀 Join Us for DataPhilly’s Tech Talks at ZeroEyes! 🚀
We will learn about Adaptive AI Systems and Decision-Centric Governance.
Our host is ZeroEyes - company that delivers a proactive, human-verified visual gun detection and situational awareness solution that integrates into existing digital security cameras to stop mass shootings and gun-related violence.
Our Gold sponsor is Liberty Personnel, widely recognized as the finest direct placement and contract recruiting firm in the region.

Event Schedule:
Doors opens at 6:00 pm ET
6:00 - 6:30 Event start and networking, DataPhilly intro
6:30 - 7:15 Johah H. Harris, Learning to Ask: Online Policy Optimization in Interactive Onboarding Systems, followed by Q&A
7:15 - 8:00 Brian M. Green, From Data Lineage to Decision Lineage: Building Infrastructure for Proportional AI Governance, followed by Q&A
After 8:00 Networking time

Speakers:
Jonah H. Harris, CEO of NEXTGRES
Learning to Ask: Online Policy Optimization in Interactive Onboarding Systems.
Abstract: Today, product managers create static, one-size-fits-most onboarding and purchase flows by hand. Every user is treated the same, locked into a fixed script that can't learn, adapt, or improve. The combinatorial complexity of possible flow permutations makes A/B testing impossible, forcing companies to optimize for a fictional "average user," which increases dropout and kills conversion.
This talk examines a real-world online learning framework created to solve this problem. It reframes onboarding as a sequential decision process. Each question or step is chosen based on prior user responses, allowing the system to adapt in real time, learn efficient macro-level flow structures, and continuously optimize micro-level question variants. By combining reinforcement learning, off-policy evaluation, and latent-state inference, this approach delivers a dynamic, personalized onboarding experience that improves with every user interaction, turning onboarding from a static artifact into a self-optimizing policy.
Bio: Jonah is a recognized expert in AI/ML, database internals, and large-scale systems, known for transforming complex systems into simple, scalable solutions. Over 25+ years, he's led startup and public technology teams, and built platforms serving hundreds of millions of users. As CEO of NEXTGRES, he is building a converged, multi-model data platform for personalization at scale. He previously served as CTO at The Meet Group and MariaDB, led AI/ML at Noom and The Meet Group, and was a founding engineer at EnterpriseDB.

Brian M. Green, Chief AI Officer & Founder, Health-Vision AI, LLC
From Data Lineage to Decision Lineage: Building Infrastructure for Proportional AI Governance.
Abstract: AI governance can fail when we treat all systems in the same manner. This talk introduces a six-archetype framework that matches AI systems to risk patterns, from low-impact internal tools to autonomous high-risk systems, and shows how decision lineage infrastructure can enable proportional controls. Decision lineage isn't just data logs, it is governance as infrastructure that makes kill switches, human escalation, and rollback mechanisms actually work.
We will look at real world examples and apply: (1) A decision tree to classify your AI systems, (2) Archetype-specific control patterns, and (3) Design patterns for embedding governance into CI/CD pipelines.
Bio: Brian is building AI governance infrastructure at [AgenticVillage.net](https://agenticvillage.net/), helping organizations implement risk-tiered frameworks that scale from experimental tools to production systems. He works with teams in healthcare, finance, and enterprise to embed governance into engineering workflows.
Please note, by RSVPing this event you agree to our Code of Conduct
Looking forward to seeing you there! 🚀