Applied AI Engineer

Foresight Works Ltd•San Francisco, CA

2d•$180,000 - $230,000

About The Position

Foresight is hiring an Applied AI Engineer to build the core machine-learning engine of its platform for delivering large-scale construction projects. The role involves architecting, training, deploying, and monitoring models for risk predictions and AI recommendations, and building backend features that make complex projects with up to 100,000 activities easier to understand. The engineer will own systems from model training through production deployment and ongoing improvement, with an emphasis on accuracy, scalability, reliability, and observability. Potential areas of work include predicting project risk and its severity; identifying critical tasks from project plans; generating project timelines and scaffolding from blueprints and 3D models; estimating task completion from multimodal inputs such as camera footage, videos, and voice notes; and recommending ways to mitigate risks or improve timelines. The role also includes repeatedly evaluating models and prototyping improvements informed by current research. The engineer will partner with full-stack engineering, product, design, and customers to identify valuable problems and turn models into production features. They will own features from technical design through release and ongoing improvement, tune performance across the modeling stack, raise standards for architecture and testing, and help shape the product and technical roadmap. Foresight is a New York-headquartered company with a small, senior team and paying customers in the data-center and construction industries. Preferred locations are New York, San Francisco, and Montreal; occasional travel may be required.

Requirements

At least 3 years of experience building and shipping machine-learning models in production, with ownership of the backend.
At least 1 year of experience building and shipping VLM/LLM-based features to production.
Strong Python backend skills, using FastAPI, Django, or similar.
Experience with relational data modeling in Postgres, including schema design, indexes, query plans, migrations, and an ORM such as SQLAlchemy.
Experience shipping agentic workflows, RAG systems, and high-volume data pipelines.
Experience with asynchronous work, including queues, background jobs, retries, idempotency, and handling worker failures.
Hands-on experience with AWS, including S3, RDS, a managed queue or event service, and containers or serverless.
Writes tests and debugs production issues using logs, metrics, and traces.
Clear written communication and comfort making pragmatic trade-offs in a fast-moving small team.

Nice To Haves

Experience with workflow orchestration for multi-step jobs, such as Temporal, Step Functions, or Airflow.
Experience with legacy-to-modern migrations or products for regulated enterprise or public-sector customers, including SOC 2 or data-residency work.
Early-stage or founding-team experience.

Responsibilities

Own features from technical design through release and beyond, measuring adoption and continuing to improve them.
Work with full-stack engineers to turn new models into production features customers rely on.
Partner with product, design, and customers to identify the problems most worth solving.
Tune performance across the modeling stack, including accuracy, latency, and reliability.
Improve the team's architecture, testing, observability, and shipping practices.
Help shape the product and technical roadmap.
Build, train, deploy, and monitor models for project risk prediction and AI recommendations.
Evaluate models repeatedly and prototype methods and models to improve feature accuracy.