Senior ML Engineer, Applied Science at Stimuler
We’re looking for an ML Engineer (Applied Science) to train models for various realtime English conversational skills improvement use cases. Example projects include:
- Personalised english learning exercise recommendation system
- Build demographic specific ASRs sensitive to various accents,
- Empathy driven role play, similar to how an offline tutor would interact with a student,
- Training models for grammar/vocab/accentuated pronunciation and similar speech feedback metrics
The role involves experimenting with different training strategies, creating datasets, building task specific evals, beating our internal SOTA, deploying in production, and repeat!
At Stimuler, we’re building real-time conversational audio AI for English conversational skills improvement as a use case. We see our main competition as offline human English speaking tutors, so building things like empathy focused user experiences, student focused learning plans, and the likes - no longer remain good-to-have’s. Our goal is to build the most effective way to learn how to speak English and have better conversations in real life, via a realtime UX.
We have a task-specific small model focused approach on our AI efforts so far. Our small models beat the best of frontier models on our specific tasks, and are a fraction of the total size!
Desired experience:
- You have 4+ years of strong experience shipping ML models in production. This is an applied science role.
- You’re comfortable creating datasets from scratch, as well as building task specific evals from scratch.
- You should think in a customer-focused way, work in a tight-knit and cross-functional environment - being a team player and willing to take on whatever is best for the company
- Mentor junior MLEs, or learn from battle-tested ones. Contribute to hiring as we expand.
- Good to have: Relevant experience working with NLP, or other speech specific models