Principal Machine Learning Engineer, Vision AI
Toast · Remote, US
About this role
**Toast — Principal Machine Learning Engineer, Vision AI** Toast builds technology that helps restaurants and local businesses succeed in a digital world—operating more efficiently, increasing sales, engaging customers, and keeping employees happy. We’re investing in **Vision AI** to help operators see more of what’s happening in their restaurants and act on it. ## Responsibilities - **Architect and build the core platform**: Design the video pipeline and essential components—from capture/ingestion to storage, processing, and delivery—so the platform is secure and reliable. - **Own the services that analyze video**: Build services that process footage, run vision models, and convert outputs into observations and metrics restaurants can act on. - **Build ML systems and drive model choices**: Select and integrate vision models (including vision-language models). Build inference pipelines, evaluation, and monitoring for production model performance; partner with ML engineers as models mature. - **Build customer-facing use cases across Toast**: Bring vision-powered features to restaurant operators across Toast products and surfaces. - **Set the technical direction**: Make key architecture decisions for the Vision AI platform—building for today while keeping a clear path forward. - **Stay hands-on**: Write production code, prototype new ideas, and ship—leading through building. - **Collaborate across orgs**: Partner with teams across Toast on integrations, dependencies, and working models. - **Lead with AI-forward engineering**: Use AI coding tools and agent harnesses as part of design, build, test, and ship workflows. - **Raise the bar**: Mentor engineers; set standards for reliability, security, and privacy; make the platform easier to build on. ## Requirements - **Deep engineering experience**: 10+ years building software, including designing and shipping distributed systems and data pipelines in production at scale. - **Backend and cloud expertise**: Strong cloud experience (AWS preferred), event-driven systems (e.g., Pulsar/Kafka), storage, and APIs; Kotlin/Java (or similar) plus comfort with Python. - **Experience with ML systems and video**: Built and operated ML systems in production (inference pipelines, evaluation, monitoring). Comfortable with video (streams, codecs, tools like ffmpeg). - **Pragmatic architecture**: Track record of designing systems that fit the current stage without over-engineering. - **Zero-to-one experience**: Take ambiguous ideas from prototype to production with real customers. - **Hands-on keyboard**: Write significant production code. - **AI-forward ways of working**: Daily use of AI coding assistants/agent harnesses; experience with LLM or vision-language model APIs. - **Collaboration and communication**: Align engineers and product partners; explain tradeoffs clearly. ## This role may not be a fit if... - You want a primarily **cross-org or architecture-only** role (this is hands-on). - You’re more attached to the **solution than the problem** (they want the simplest approach that solves real operator needs). - You tend to build for **scale before it’s needed**. - You prefer **established processes/tooling** (this team shapes its own workflows). - You’re uncomfortable with **ambiguity**. - You’d rather be **perfect than experiment** (they test with real restaurants and iterate). - AI-assisted development isn’t part of your workflow. ## Special Sauce (Nice to Haves) - Computer vision, video analytics
Listing freshness
CronJobs last confirmed this listing 11h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.