Senior Researcher (Machine Learning)

Our Client is a AI start-up company. You will own the research and intelligence direction for our core system. Your mission is to define how our AI reasons, evaluates, and continuously improves within a high-frequency production environment. This is a role for a pragmatic visionary who shapes early product intelligence by blending research rigor with real-world application.

What You'll Do

• Steer Core Intelligence Direction: Define and evolve the research roadmap for core AI capabilities, including context representation, memory, reasoning, planning, and orchestration.
• Architect vs. Leverage Decision-Making: Determine when to design proprietary model architectures versus adapting and fine-tuning frontier commercial or open-source models.
• Build Pragmatic Evaluation Frameworks: Establish robust evaluation guardrails that prioritize real-world usefulness, safety, and long-term behavioral reliability over vanity benchmarks.
• Explore Frontier AI Techniques: Guide the hands-on exploration of advanced methodologies, including retrieval-augmented training, mixture-of-experts, distillation, and multi-agent systems.
• Partner on Product Strategy: Work in lockstep with product and application engineering to embed safety, alignment, and elite intelligence standards into the core product UX.

Requirements

• Production-Proven ML Expertise: Deep experience building, scaling, and evolving high-frequency machine learning systems successfully deployed in production.
• Elite Technical Judgment: Exceptional intuition regarding model behavior, failure modes, long-horizon trade-offs, and making high-impact architectural decisions with incomplete data.
• Practical Builder's Mindset: An obsession with execution, correctness, and real-world system behavior over academic publishing or incremental benchmark chasing.
• Founder-Level Ownership: A high-agency, founder-style mentality with a proven track record of setting the technical bar for research rigor and organization-wide quality.
• Advanced Infrastructure Literacy: Proficiency with Python, PyTorch/JAX, and navigating GPU-based training and inference infrastructure.