Founding AI Engineer / AI Researcher

We are the team behind Kidooo AI, the AI agent for personalized parenting and child development. We help parents understand their child’s progress, turn everyday observations into clear next steps, and build more confident daily routines. We want an AI engineer who can take research ideas all the way to production, and who understands that in this domain a confident wrong answer is worse than no answer at all.

• The Kidooo agent: Design the reasoning, memory, and personalization layer that turns a parent’s everyday observations into clear, developmentally appropriate next steps.
• Multimodal understanding: Build pipelines that interpret photos, short videos, voice notes, and free-text logs of real family moments, and extract useful developmental signal from messy input.
• Grounding in child development: Work with our domain advisors to anchor model outputs in established developmental frameworks and evidence, so guidance is trustworthy rather than merely fluent.
• Evaluation infrastructure: Stand up the eval sets, regression suites, and human-review loops that tell us whether a change actually made Kidooo better. You will own quality, not just capability.
• Safety and privacy: Build guardrails for a health-adjacent product used by families, including refusal and escalation behavior, bias testing across families and cultures, and privacy-preserving data handling.
• Cost and latency: Make it fast and affordable at scale through model routing, caching, distillation, and knowing when a small model is the right answer.

• Technically deep: BS, MS, or PhD in Computer Science, Machine Learning, or a related technical field from a top-tier university, or equivalent demonstrated ability. Publications, an active GitHub, or shipped AI products all count.
• 2+ years building AI systems in production, including hands-on work with LLMs: prompting, retrieval, agents, fine-tuning, and evaluation. Strong Python required.
• Research-to-product bridge: You read the literature and can implement it, but you optimize for what reaches parents this month rather than what would look best in a paper.
• Rigorous about evaluation: You are suspicious of demos. You want the numbers, the failure cases, and the edge cases before you call something done.
• Serious about safety: You understand that families and children are a sensitive domain, and you build with humility, appropriate uncertainty, and clear boundaries around when to defer to a professional.
• Mission driven: You care deeply about parenting, child development, and practical family support, and want your work to have real social impact.
• Self-directed: You thrive in an early-stage environment where you own the problem, not just the model.

Nice to have: experience in health, education, or other regulated or high-trust domains; background in developmental psychology or cognitive science; multimodal or speech modeling; on-device and edge inference.