About the role
Forgetwell seeks a focused role to improve retrieval quality for its capture-first note app. You will enhance exact and meaning-based recall, embeddings, and multilingual search to ensure precise results from short fragments and user queries.
You will work on privacy-preserving ML, on-device inference where appropriate, and latency-conscious design to keep the experience fast and private at scale.
Own end-to-end retrieval quality: models, chunking, ranking and evaluation. Make recall understand intent, including time ranges, entities, and questions. Build evaluation sets from messy notes and measure retrieval quality. Improve multilingual search and meaning-based recall on real vocabularies. Keep inference fast and affordable at single-binary scale. Protect capture speed and the calm product experience as retrieval grows. Shipped retrieval or embedding systems to production. Strong privacy-preserving ML stance: data minimization, on-device processing. Experience evaluating retrieval on real-world noisy text. Ability to measure before tuning and avoid vibes-based quality claims.Skills
Interested in this role?
You'll apply directly on Forgetwell's careers site.