Lead Data Scientist / Machine Learning Engineer – Knowledge Graph & Agentic AI

About River River is the human layer of the internet. The consent and ownership infrastructure that puts people back in the loop. Users see their data, control who accesses it, earn 70% of the value it creates, and get AI that finally serves them, not the platform. We are pre-revenue, raising a $3M seed at a $30M pre-money, and preparing for market launch. Our founding CEO brought MTV and NetJets to Europe and we are backed by technology luminaries from Google, Sun Microsystems, and beyond. We are looking for a Lead Data Scientist to take full ownership of everything data science at River – from our personal data graph to our next-generation agentic AI systems. This is a foundational hire that will underpin our expanding team. You will be the person who defines how River understands, connects, and reasons over user data at scale, and who builds the data science function from the ground up. What We're Building River is three products, one platform: • River Social – live and growing. A social platform where you own your data, control your identity, and get paid when it's used. 70% of value goes back to you. • River Source – enterprise consent infrastructure. The bridge that gives AI platforms and brands access to high-fidelity, consented user data at scale. • RiverAI – launching 2026. A desktop AI client built on your personal data graph, with real agentic capabilities that work for you, not just talk to you. RiverAI is powered by Rivera, River's AI engine, which is unique: • It knows you. Rivera reasons over your personal data graph – your preferences, relationships, history, and context – to deliver intelligence that's genuinely yours. • It's agentic. Rivera doesn't just answer questions – it takes action on your behalf, orchestrating multi-step workflows across your data and services with full transparency and user consent at every step. • It's proactive and predictive. Rivera anticipates your needs and offers to help without you even asking. • It's portable. Your data identity moves with you. Rivera works across River's platform and beyond, acting with your authority wherever AI meets the individual. What You'll Be Doing • Designing and evolving River's personal data graph – defining entity schemas, relationship models, and enrichment pipelines that transform raw user data from diverse platforms into a rich, interconnected graph of individual intelligence. • Building and maintaining data taxonomies – creating structured classification systems that organize the vast diversity of user data into coherent, queryable categories, enabling consistent data interpretation across River Social, River Source, and RiverAI. • Cleaning, normalizing, and structuring noisy data – building robust pipelines that ingest messy, heterogeneous real-world data from multiple platforms and sources, resolving inconsistencies, deduplicating entities, and transforming it into high-fidelity graph-ready data at scale. • Advancing natural language processing capabilities – developing NLP systems that extract meaning, intent, and relationships from unstructured text data, powering Rivera's understanding of user context and enabling semantic search across the personal data graph. • Building agentic AI systems for RiverAI – architecting multi-agent frameworks where Rivera can reason over the personal data graph, plan multi-step actions, and execute autonomous workflows on behalf of users. • Staying at the cutting edge of AI development – continuously evaluating and integrating emerging techniques in machine learning, graph ML, embedding models, and agentic architectures to ensure River's AI capabilities remain state-of-the-art. • Developing AI-powered data ingestion for River Source – building intelligent pipelines that automatically analyze, classify, and enrich enterprise and user data as it enters the graph, powering River's consent infrastructure and maximizing network effects. • Creating novel data valuation metrics – quantifying the value of user data contributions within the graph to power River Social's compensation model, where 70% of value flows back to users. • Designing reinforcement learning systems for Rivera's personalized recommendation engine, leveraging graph structure to deliver context-aware, relationship-informed suggestions. • Implementing hybrid search architectures (vector + graph traversal + traditional) with real-time trend analysis across the personal data graph. • Building multi-agent preference aggregation models for group commerce features, enabling Rivera to reason over the preferences of multiple connected users simultaneously. • Creating explainable AI components that reveal decision-making processes, ensuring users understand why Rivera takes actions and how their data informs recommendations. • Defining and measuring success – establishing evaluation frameworks, A/B testing infrastructure, and metrics for graph quality, agent reliability, and user satisfaction at scale.