New Grad Machine Learning Engineer

## About the role OtterHalf is a dating app for South Asian singles in the U.S. and Canada. We are building a more intentional alternative to endless swiping, with prompt-led profiles, clearer likes, and a product experience designed around better conversations. We are looking for a New Grad Machine Learning Engineer to help improve recommendation quality, ranking, profile understanding, trust signals, and data-informed product systems. This role is not about building ML for the sake of ML. The goal is to use data and practical modeling to make the product more relevant, respectful, safe, and useful for members. ## What you'll work on - Explore recommendation and ranking approaches for discovery surfaces - Analyze profile completion, likes, matches, engagement, and conversation-start behavior - Help define product-quality metrics for matching and discovery - Prototype lightweight models or scoring systems for profile quality, relevance, and trust - Explore text, profile, and interest understanding in a privacy-conscious way - Partner with product and moderation workstreams on trust and safety signals - Evaluate experiments clearly and avoid overcomplicating the product - Write clear analysis docs and explain tradeoffs to non-ML stakeholders ## Requirements - Recent graduate or early-career engineer with strong fundamentals in machine learning, data science, statistics, or applied math - Strong programming ability in a language commonly used for data or ML work - Comfortable with data analysis, experimentation, model evaluation, and clear technical writing - Familiarity with ranking, recommendations, classification, embeddings, NLP, or user-behavior modeling - Ability to balance model quality, privacy, fairness, latency, and product simplicity - Comfortable working with messy product data and imperfect early-stage datasets - Strong judgment about when a simple heuristic is better than a complex model - Interest in consumer apps, dating apps, social products, trust systems, or recommendations ## Nice to have - Experience building ML, analytics, or data systems beyond coursework-only examples - Experience with recommendation systems, search, ranking, NLP, embeddings, or classification - Experience analyzing consumer product funnels or engagement behavior - Experience with trust and safety, moderation systems, spam detection, or profile quality signals - Experience writing clear experiment summaries or model evaluation reports ## Who this is not a fit for - Someone who wants only research work with no product constraints - Someone who wants to train large models without caring about product impact - Someone who cannot explain model tradeoffs clearly and practically ## How to apply Please include your resume or LinkedIn profile, links to relevant projects if available, and a short note about an ML or data project where you had to make a practical tradeoff.