SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Sandbar is a New York-based interface company building hardware and software products to augment human cognition. The company's first product, Stream, is a private voice ring and conversational interface launching in Summer 2026, with coverage in WSJ, Bloomberg, and Wired.
You will join as ML Technical Lead to design and lead the agentic interactions team within Stream. This is a hands-on leadership role focused on consumer product—not research or platform infrastructure. You will set technical direction for ML-based features, design user-facing model experiences including agentic control and conversational interactions, and own the end-to-end quality of the agent in customers' hands.
Key responsibilities include:
- Lead architecture, strategy, and technical tradeoffs for ML-based products
- Design and improve user-facing model experiences, including agentic control, conversational interactions, and information retrieval
- Develop infrastructure for rapid model development, evaluation, and deployment
- Optimize inference pipelines for latency, reliability, and concurrency
- Collaborate closely with ML/AI engineers, infrastructure engineers, designers, and cofounders on existing and future products
You will work directly with the founding team and report into the broader engineering organization. The role emphasizes hands-on technical contribution alongside leadership—you will both set direction and build.
Qualifications:
- 6+ years of machine learning experience
- Proven track record shipping ML-based products (required)
- Experience in large-language model evaluation and post-training (required)
- Experience in audio/voice ML (preferred)
- Comfortable writing production-grade software
- Passion for human-computer interaction
- Experience setting technical direction and driving execution with multidisciplinary teams
