AMA: Cortex VP of Product, Kara Gillis on Product Development Process

Our product development process starts with tiered feature classification (Tier 1: flagship, Tier 2: new functionality, Tier 3: enhancements) that determines the rigor of validation required, then moves through two distinct review gates—Problem Definition Review to align on customer jobs-to-be-done and requirements with a small group including tech leads, followed by Product Definition Review with the implementation team to validate solution approach and technical feasibility using a living PRD ...Read More

We use a tiered approach that matches the validation rigor to the feature's business impact: Participant Selection: Tier 1 features (flagship/new products): Start with design partners, building to 5+ customers in private beta and 7-10+ by public beta Tier 2 features (new functionality): 3+ customers in private beta, scaling to 3-5+ by public beta Tier 3 features (enhancements): Generally skip beta and go direct to GA Participants are selected based on their use case fit and willingness to provid ...Read More

We use a two-stage review process that separates "are we solving the right problem?" from "are we solving the problem right?" This ensures engineering understands not just the technical scope, but the customer context and constraints before diving into implementation. Stage 1: Problem Definition Review Purpose: Align on customer requirements before discussing solutions Participants: PMs, Design, Executive Leadership, Designated Tech Lead What Engineering Learns: Problem statement: What customer ...Read More

Based on my experience at Cortex, here's how I'd approach adapting the development process for ML/AI features: All AI-driven features go through a Research Preview designation before GA. This serves two critical purposes: Sales enablement: Clearly flags "AI" features so prospects and AI-skeptic customers can opt-in, avoiding lengthy security reviews and contract addendums Risk management: Keeps features behind feature flags until they meet quality thresholds This acknowledges upfront that AI fea ...Read More

I'm a big fan of quarterly planning. But I absolutely HATE overly complicated, unnecessarily long planning of any kind. I like the kind of planning that provides clarity, cuts down on chaos, but leaves enough flexibility to pivot if necessary. I'll tell you what I implemented at Cortex in the last month. We moved from a six month high level planning process to a more precise three month cadence. Quarters are the perfect amount of time to churn out a ton of PRDs (and ERDs) to fill up the next 1-2 ...Read More

Our post-launch review process is embedded directly into our stage gate criteria and living documentation approach, creating continuous feedback loops rather than discrete post-mortems. For features to exit Public Beta and reach GA, we require structured validation that doubles as our initial post-launch review: Tier 1 features: 3+ product or CX-led customer interviews confirming use cases are being achieved, plus a design partner willing to provide public testimonials (press quotes, case studie ...Read More