Founding Machine Learning Engineer
AI agents are taking the world by storm–doing everything from booking reservations to ordering things online. Despite this, AI lacks the real world context that drives some of our most important decisions. Millions of conversations, thoughts, and crucial decisions are made in the real world every day–at a work meeting, on a walk, or in a warehouse.
Today, all of that valuable context is instantly thrown away. But we're building a future where AI understands what you're doing in real time, solves problems for you, and helps you think, before you even have to ask. Today, MIRA serves thousands of real conversations across courtrooms, machine shops, police departments, and warehouses.
You would lead the development of a first-of-its-kind model that proactively fact-checks claims, generates tasks, and recalls information, and deploy it to real, crucial environments where decisions actually change lives.
We recently sold out and shipped our first batch of units, and we're moving fast. MIRA is backed by General Catalyst, Naval Ravikant, Pillar VC, and more.
Responsibilities
- Build and ship machine learning systems that turn conversations and ambient context into useful, proactive insights.
- Develop and evaluate retrieval, ranking, and context selection so the right information reaches the model at the right time.
- Work on real-time audio and language model workflows, balancing response quality, latency, reliability, and cost.
- Build datasets, evaluation methods, and feedback loops to measure and improve memory recall and proactive behavior
- Partner with software, firmware, and product teammates to take experiments into production and improve the experience on real devices.
- Design with user privacy and appropriate handling of personal context in mind.
Qualifications
- 3+ years of experience building machine learning systems.
- Strong programming and engineering fundamentals, with experience taking systems from prototype to production.
- Ability to work through ambiguous problems, run practical experiments, and make sound product and engineering tradeoffs.
- This role is full-time, in-person in San Francisco.
These might make you a great fit
- Experience with multimodal models, speech or audio, fine-tuning, or low-latency inference.
- Experience building memory, personalization, or agent systems.
- Prior experience at an early-stage startup or deploying consumer AI experiences.
