ML Engineer

Visia; New York (NY); On-site; USD 140,000 - 155,000; Full time

Benefits offered by this job

Flexible working hours

Unlimited PTO

Equity compensation

Job summary

A leading AI technology company in New York seeks a Platform Engineer to own and accelerate their data flywheel. This role involves building tooling and automation for annotation, model training, and data pipelines. Ideal candidates have strong programming skills in Python, TypeScript, or Go, and are familiar with cloud computing and CI/CD. The compensation ranges from $140,000 to $155,000 annually with additional equity, alongside flexible working conditions and unlimited PTO.

Qualifications

• Strong programming skills in Python, TypeScript, or Go.
• Experience with frontend development, cloud computing, and CI/CD.
• Willingness to learn about ML infrastructure.

Responsibilities

• Automate instance-segmentation and vision-language model annotation tooling.
• Build and maintain data pipelines for model training.
• Develop tooling for model performance monitoring.

Skills

Programming skills in Python

Programming skills in TypeScript

Programming skills in Go

Frontend development (React / React Native)

Cloud computing (GCP, AWS, or Azure)

CI/CD, IaC, Docker, Terraform

Tools

GitHub Actions

Computer vision pipelines

Job description

Visia is the first multimodal AI platform custom built for heavy industry. Visia’s full-stack physical intelligence platform includes robust sensing systems across imaging modes (cameras, X-rays, cargo X-rays, LiDAR), foundation vision-language models that convert raw sensor data into structured operational intelligence, and software-driven Field Engineering that drives real transformation on-site at some of the world’s largest industrial operations.

We deploy our hardware systems powered by our model flywheel across recycling facilities, steel mills, aluminum smelters, ports, and waste-to-energy plants, processing over 1 billion data points a year. When customers adopt Visia they aren’t buying a point solution – they are parting with a platform business that turns unstructured optical data into real-time, actionable intelligence for their operations.

What We're Looking For

We're looking for a Platform Engineer to own and accelerate Visia's data flywheel — the engine that makes our fine-tuned models smarter with every deployment. You'll primarily build the tooling and automation that powers our annotation, model training, and data pipelines, ensuring that every image we process makes Solstice better at understanding the physical world.

This is a high-leverage role. The systems you build directly determine how fast we can onboard new customers, new sensor modalities, and new industry verticals. You'll also pitch in on custom Field Engineering implementations when needed and contribute to internal agentic tooling that helps the team move faster.

Responsibilities

• Data Flywheel & Model Tooling
  • Automate and improve our instance-segmentation and vision-language model annotation tooling, creating smooth, efficient workflows for our annotators and model managers
  • Build and maintain the pipelines that move data from deployed sensors → annotation → model training → production deployment
  • Develop tooling for model performance monitoring, error analysis, and continuous improvement across customer sites
  • Work closely with ML engineers to accelerate the feedback loop between deployed models and training infrastructure
• Contribute to custom Visia implementations for enterprise customers — building integrations with scale systems, ERPs, PLCs, and customer-specific reporting
• Support Forward Deployed Engineering (FDE) engagements by building reusable tooling and templates that make each deployment faster than the last
• Help translate customer operational needs into technical specifications and working softwareInternal
• Tooling & Automation
  • Build agentic tooling to accelerate GTM workflows — automated sales deck generation, customer research, and outreach
  • Collaborate with the CEO on automating operational workflows
  • Contribute to Visia's overall product roadmap and help shape the future of physical intelligence in heavy industry

Who You Are

• You have strong programming skills in one or more commonly used programming languages (we use Python, TypeScript, and Go)
• You move fast, ship iteratively, and care about building things that actually work in production
• You're comfortable working with at least some of the following and eager to learn the rest:
• Frontend development (React / React Native)
• Cloud computing (GCP, AWS, or Azure)
• CI/CD, IaC, and containerization (Docker, Terraform, GitHub Actions)
• Bonus: experience with computer vision pipelines, annotation tooling, or ML infrastructure
• You have proof of your engineering ability through GitHub projects or work experience

What You Might Work On

• Building automated annotation pipelines that use SAM and vision-language models to pre-label datasets, cutting model iteration cycles by 5-10x
• Creating model performance dashboards that surface accuracy regressions across customer sites before they become customer issues
• Integrating Visia with a customer's scale system to automate compliance reporting they currently do by hand
• Building an internal agent that generates customized sales decks from customer research and Visia's deployment playbooks
• Compensation: $140,000-$155,000/yr + ~$30,000/yr in equity ($170,000-$185,000/yr)
• Flexible + Unlimited PTO