Director / Principal Product Manager – AGI Reasoning Platform

JazzX AI•San Francisco, CA

About The Position

About SAIGroup SAIGroup is a private investment firm that has committed $1 billion to incubate and scale revolutionary AI-powered enterprise software application companies. Our portfolio, a testament to our success, comprises rapidly growing AI companies that collectively cater to over 2,000+ major global customers, approaching $800 million in annual revenue, and employing a global workforce of over 4,000 individuals. SAIGroup invests in new ventures based on breakthrough AI-based products that have the potential to disrupt existing enterprise software markets. SAIGroup’s latest investment, JazzX AI , is a pioneering technology company on a mission to shape the future of work through an AGI platform purpose-built for the enterprise. JazzX AI is not just building another AI tool—it’s reimagining business processes from the ground up, enabling seamless collaboration between humans and intelligent systems. The result is a dramatic leap in productivity, efficiency, and decision velocity, empowering enterprises to become pacesetters who lead their industries and set new benchmarks for innovation and excellence. Role Overview This role leads the productization of the Reasoning Platform driving definition and delivery of the core reasoning capabilities underpinning the JazzX AGI system. This covers a wide spectrum of reasoning approaches from neuro-symbolic, probabilistic, cognitive, reinforcement-learning (RL)-based reasoning systems, as well as the infrastructure, tooling, and evaluation frameworks required to build, train, validate, and continuously improve AGI reasoners. The role owns strategy across data curation, benchmarking, feature alignment with customer use cases, RL gym environments, safety constraints, and tools enabling forward deployment teams to design, evaluate, and refine reasoning behaviors at scale. The position works closely with research, engineering, safety, design, forward-deployment teams, and enterprise domain experts to ensure reasoning is predictable, verifiable, explainable, and aligned with real-world needs.

Requirements

10+ years in technical product management; 5+ years in AI/ML, agent systems, RL, or reasoning technologies.
Experience building large-scale AI platforms, RL training systems, evaluation frameworks, or advanced decision-support tools.
Demonstrated success in productizing research-heavy technologies.
Strong understanding of RL, policy optimization, reward modeling, agentic planning, and model fine tuning approaches.
Knowledge of evaluation methods, benchmarking design, dataset development, and safety frameworks.
Ability to lead multi-year product strategy across research and engineering.
Skilled at turning emerging reasoning research into stable, scalable platform features.
Ability to mentor senior PMs across data curation, model-based reasoning, and builder experiences.
Bachelor’s degree in Computer Science, Engineering, or a related technical field required.

Nice To Haves

Advanced degrees (MS, PhD) in relevant areas preferred but not required.

Responsibilities

Product Strategy Define the roadmap for the Reasoner Platform spanning the full spectrum of reasoning systems. Set product direction for prioritized feature development aligned with customer use cases and value drivers. Drive strategy for evaluation, safety constraints, reward models, and benchmarking systems improving reasoning accuracy, stability and reliability.
Reasoner Platform Productization Define requirements for configurable RL gym environments used to train, test, and validate agentic behaviors. Convert advanced reasoning research into productized capabilities, APIs, and developer tools for forward deployment teams to build customer solutions. Partner with research and engineering to deliver scalable training and inference infrastructure for reasoning-centric workloads.
Data Curation & Benchmarking Lead strategy for datasets, scenarios, and evaluation frameworks that measure reasoning accuracy, robustness, alignment, and generalization. Define requirements for synthetic data generation, curated reasoning traces, and adversarial testing. Own the roadmap for benchmarking tools, dashboards, leaderboards, and automated evaluation pipelines.
Reasoner Builder & Developer Experience Deliver tools that enable teams to design, test, refine, and deploy custom reasoning patterns. Define interfaces for rule editing, prompt program construction, policy tuning, and RL reward configuration. Ensure builder workflows support rapid iteration, safe experimentation, and interpretable outputs.
Integration with AGI Capabilities Align reasoning capabilities with memory, grounding, planning, and the broader Knowledge Platform. Define safe mechanisms for reasoners to inspect state, update plans, call tools, interact with environments, and refine policies. Integrate provenance, safety checks, transparency, and control mechanisms across all reasoning operations.
Customer & Domain Alignment Partner with enterprise domain specialists to understand complex decision workflows and map them to reasoning primitives. Identify common reasoning patterns across industries and convert them into reusable platform components. Ensure the platform supports both autonomous agents and human-in-the-loop decision processes.
Metrics & Impact Define and track core reasoning KPIs such as correctness, robustness, consistency, sample efficiency, safety adherence, and generalization. Measure performance across synthetic, human-generated, and real-world tasks. Use evaluation insights to refine roadmap priorities and ensure platform impact translates to real-world performance.