WHY YOU SHOULD JOIN US

At Solstice, we're building AI software that helps life sciences teams turn complex scientific and brand information into marketing content they can create, refine, and review with confidence. We've worked with leading life sciences manufacturers and more than 50 pharma brands. You'll build the systems and experiences that make this work faster and easier.

Build the foundation for how people work with AI. Your work will span the infrastructure that runs agents and the product experiences that help customers direct, inspect, and use their output.

Take on engineering problems with depth. Make long-running agents reliable, stream progress into responsive interfaces, deliver the right context securely, and understand what happens across every step of execution.

Have a direct hand in what we build. Join an early team where you'll shape architecture, make product tradeoffs, and work closely with our co-founders, Yiwen and Aris. You'll see your decisions reach customers and use their feedback to improve the product.

We're backed by Transformation Capital, Twelve Below, Virtue, and the founders of Datavant, Commure, and Paradigm.

SOLSTICE IN TORONTO

We're opening our first engineering office outside New York, in Toronto, and you'll be one of its first engineers. The Toronto team will own a product area end to end, with its own roadmap and on-call. It won't be a queue for work handed over from New York.

Same time zone as New York, and one of our co-founders will be in Toronto with the team.

You'll spend your first month in New York working alongside the team in person. We cover travel and accommodation.

This role will be on-site in Toronto. We'll start remotely for the first few months, with a WeWork stipend for those who prefer it. As the team grows, we'll move into our own Toronto office by the end of the year and work from there five days a week.

ABOUT THE ROLE

We're looking for a Fullstack AI Engineer to build and maintain our agent platform and bring its capabilities into the product. You'll own work across backend services, APIs, agent orchestration, and customer-facing interfaces, with a strong focus on Python and platform engineering.

Python is our primary backend language, and FastAPI powers our APIs. You'll help decide how agents run, recover from failures, access tools and context, and communicate with the application. You'll also build the React and Next.js experiences that give users visibility and control over that work.

You'll partner with product, design, Applied AI, and the broader engineering team to take problems from an initial idea through architecture, implementation, and production. The scope includes both building new capabilities and making the systems behind them easier to operate and extend.

WHAT YOU'LL BE WORKING ON

Build and maintain the agent platform. Use Python, FastAPI, Restate, and Redis to build the shared systems that power our agents. Shape session management, orchestration, tool access, and context delivery so new AI capabilities can build on a dependable foundation.

Design APIs that power the product. Build Python and FastAPI services for content creation, document processing, collaboration, and enterprise integrations. Design clear API contracts and PostgreSQL data models that make complex workflows straightforward to build on.

Solve the hard parts of agent execution. Make long-running workflows recover from failures, handle cancellation and retries, and run in isolated environments. Give agents current documents, assets, and brand context while keeping customer data and permissions properly scoped.

Bring AI workflows to life in the product. Build with React, Next.js, and TypeScript to show streaming progress and tool activity, support reconnectable sessions, and give users intuitive ways to review, accept, or refine AI output.

Make the whole system observable. Connect APIs, orchestration, and agent execution through Datadog and OpenTelemetry tracing, structured logs, metrics, and alerts. Build the visibility to explain failures, find bottlenecks, and improve latency, reliability, and cost.