Richard Wolff
Enterprise AI leader building multi-agent systems and governed AI workflows for global-scale platforms.
Plano, Texas, United States
Actions
Richard Wolff is a Director of Technology at PwC focused on AI engineering, global architecture, and production delivery for enterprise platforms. His work centers on making AI useful inside large, data-sensitive organizations: governed agent workflows, reliable knowledge access, workflow automation, integration architecture, and operating models that can scale beyond demos.
Richard's recent public writing and hands-on work focus on practical agentic systems: turning one-off AI work into reusable tools, using eval-driven design to separate flaky paths from wrong paths, improving signal-to-noise in knowledge work, preserving human tone in AI-assisted communication, and combining deterministic checks with model-driven judgment.
He brings a practitioner perspective for audiences working through the next phase of enterprise AI adoption: how to design agent workflows that are trustworthy, observable, reusable, and valuable at the workflow level.
Links
Badges
Area of Expertise
Topics
From RAG to Workflows to Agents: Choosing the Right Architecture
Enterprise teams are often told that agents are the next step after retrieval. That shortcut creates expensive systems that are difficult to test, govern, and operate. This session offers a practical decision framework for choosing the simplest architecture that can safely deliver the outcome.
We will compare three patterns through the same enterprise workflow: retrieval when the task is primarily about finding trustworthy context; deterministic workflows when rules, approvals, and integrations should control the path; and tool-using agents when the work genuinely requires judgment across changing context. For each pattern, we will examine failure modes, operational signals, evaluation, human review, and the point at which multi-agent coordination creates more cost and risk than value.
The goal is not to make every workflow agentic. It is to make architecture choices explicit, observable, and reversible. Attendees will leave with a decision framework they can use in design reviews, plus practical questions for data boundaries, tool permissions, evaluation, observability, retries, and escalation before an AI system reaches production.
Beyond the AI Pilot: Operating Enterprise AI Platforms at Scale
Most enterprise AI programs do not fail because the model is weak. They stall because teams cannot operate AI-enabled workflows with the same discipline they expect from production software: clear ownership, controlled releases, observable behavior, reliable fallback paths, and feedback that improves the system over time.
This session presents an operating model for moving from isolated AI pilots to a dependable enterprise platform. We will walk through the platform capabilities that make AI workloads sustainable: reusable integration and context patterns, deterministic controls around model-driven judgment, release and evaluation gates, telemetry for quality, latency, cost, and policy signals, human escalation for uncertain outcomes, and incident learning that feeds the next delivery cycle.
The emphasis is not on a vendor stack or a single autonomous-agent pattern. It is on the engineering choices that let teams scale AI responsibly: where to standardize, what to observe, how to preserve accountability, and how to make reliability a delivery habit rather than a late-stage rescue effort. Attendees will leave with a practical scorecard for assessing whether an AI initiative is ready to move from pilot to operated platform.
Shift-Left Governance for Faster Agentic AI Delivery
Agentic AI in financial services will not move from experiments to production because a model got smarter. It will move when firms can prove how the workflow was built, what data it touched, which tools it invoked, what controls were tested, and why it was safe to release.
This session presents a delivery-cycle governance harness for agentic AI: codified risk objectives, deterministic tests, eval-driven controls, workload identity, data lineage, provenance, policy gates, and runtime evidence. The goal is to shift governance left so compliance becomes engineering rigor, not a questionnaire at the end.
The talk is practical. Reinforcement learning, prompt tuning, and eval scores can improve behavior, but they are not governance. They do not prove identity, lineage, authorization, auditability, or release readiness.
We will walk through how financial-services teams can define risk objectives early, turn them into tests and gates, establish trusted workload identity with standards such as SPIFFE/SPIRE, require data lineage and provenance, and operate live agentic workflows with runtime evidence. The result is faster delivery because trust is built in early, not inspected in late.
Event-Driven Agents: Inference Architecture at Production Scale
A work item lands on a queue. It might be a document to classify, a case to route, a knowledge update to inspect, or an exception that needs review. At first, it is tempting to treat that as a simple model call: send the payload to an LLM, get a result, and move on. At production scale, that decision is rarely so simple.
This session focuses on inference architecture for event-driven agents: when inference should run, what context it needs, what result event it should emit, and how the workflow should behave when the model is uncertain or wrong. We will walk through a practical pattern: deterministic pre-checks validate the event, inference runs only when model use is justified, results emit new events with confidence and trace metadata, low-confidence outcomes route to human review, and failures land in dead-letter flows that become operational and governance signals.
From there, we will connect the pieces into a production architecture: queues for discrete inference work, streams and topics for multi-consumer state changes, agent memory or vector indexes as derived context, and warehouses or lakehouses as the feedback layer for evaluation, drift review, audit, and cost tracking. The patterns apply across streaming-backed architectures where events carry context, trigger inference, and preserve a record of what happened.
The core question is not whether agents should use events. It is how events should control inference: what should trigger a model call, what should stay deterministic, what should be persisted or replayed, what should become evaluation data, and where a human needs to stay in the loop. Attendees will leave with a practical framework for designing event-driven inference workflows that are observable, reliable, and governable in production.
Richard Wolff
Enterprise AI leader building multi-agent systems and governed AI workflows for global-scale platforms.
Plano, Texas, United States
Links
Actions
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top