Turning complex technology into products that matter.

As Product DRI for Duo Developer, I worked with Engineering to distinguish model limitations from context, tool, and execution issues. I made the prioritization call for compaction after reviewing the evidence, rather than relying only on more available context or future model improvements. The trade-off was investing in shared context management while protecting information that longer coding tasks depend on.

I tracked task completion weekly and used SWE-bench Pro and Verified alongside latency, costs to us and customers, production telemetry, and developer feedback. Real repositories and execution environments could reveal problems that standardized benchmarks missed.

My focus was the product judgment behind the investment: where shared infrastructure would improve developer outcomes, and how to keep learning as models and production needs evolved.

I evaluated compaction frequency and cost when deciding how to support it across deployment models, prioritizing a consistent implementation over deployment-specific forks. This connected agent reliability with operating cost and implementation complexity.

I influenced the direction toward token-based or hybrid billing rather than pure request-based pricing. My contribution connected model selection and task routing with quality, latency, and inference economics. A request alone could hide substantial differences in the work an agent performed; the pricing discussion needed to account for that variability.

As product DRI, I partnered with a solutions consultant to build and validate PD Live with paying enterprise customers. I then led the business case and financial model to secure investment and bring the concept into PagerDuty’s AIOps Operations Console, a centralized experience helping operations teams monitor incidents, prioritize signals, and identify remediation actions.

I led the launch of the consumption-based AIOps SKU, owning monetization strategy, billing readiness, and go-to-market execution. I partnered with Solutions Engineering and Customer Success to translate the offering into enterprise value conversations and adoption plans.