Gregory Maulick
Product Manager — AI Products & Platforms
This site showcases my work and product decisions across AI products, including customer-facing applications and live experiments in how AI systems are designed, evaluated, and improved for real users.
Explore My Work
I builds AI products that solve real customer problems and scale to global usage
- Directed product strategy for a 40-engineer cross-functional organization, shipping systems used by 3B+ users.
- Defined the roadmap for a 0→1 platform, scaling it from concept to $40M ARR.
Ask the Gregory Agent
A purpose-built LLM augmented with a RAG pipeline over Gregory's professional and personal history, designed to evaluate and improve response quality. The long-term goal is to iteratively refine the system to achieve consistent 90th-percentile performance
- Walk me through Gregory's biggest 0 to 1 product
- How does Gregory balance model quality with user experience
Experience
Owned the AI platform transforming live match moments into automated video and contextual advertising and monetization at global scale. Defined the commercial model, set platform strategy across automation, scalability, reliability, and cost, and institutionalized experimentation to drive sustained revenue expansion.
Operationalized experimental computer vision models into production AI workflows deployed across global enterprise customers. Set evaluation thresholds, enforced release gates, and established rollout discipline for reliable scale.
Drove strategic shift from research positioning to commercial growth, redefining product narrative, onboarding experience, and roadmap priorities around buyer use cases and revenue impact.
Owned acquisition and activation strategy, instrumenting funnels and simplifying UX to accelerate first-value and reduce friction across developer workflows.
Projects
Chatbot Metrics & Observability
The AI assistant available at the top of this page is part of a hands-on project I built to deepen my understanding of how LLM-based products are designed, evaluated, and iterated in real-world settings. The system combines a lightweight LLM with a RAG pipeline grounded in my professional and personal history, and is intentionally instrumented with evaluation metrics such as faithfulness, relevance, and naturalness. These metrics help surface failure modes, assess response quality, and guide iterative improvements—reflecting the responsibilities of a product manager working on LLM-powered systems. The long-term goal of the project is to continuously refine the system toward consistent 90th-percentile performance, while developing stronger intuition around the tradeoffs between retrieval, generation, and user experience.
Health.AI (Featured on TEDx)
Health.AI is an innovative Android application designed to detect melanoma, leveraging advanced artificial intelligence technology. The app aims to reduce skin cancer deaths in under-resourced regions by providing accessible, accurate skin assessments directly to users. By utilizing computer vision algorithms, Health.AI analyzes images of skin lesions to identify potential malignancies, enabling early intervention and treatment. The application gained recognition for its impactful potential, culminating in a feature on TEDx, which highlights its significance in promoting health equity and improving outcomes in underserved communities. Health.AI exemplifies the intersection of technology and healthcare, showcasing how AI can be harnessed to address critical public health challenges.