Member of Technical Staff - Fullstack AI Engineer

Solstice is hiring a Fullstack AI Engineer in New York, NY (onsite) to build and maintain an agent platform and bring agent capabilities into the product. This role connects backend services and APIs, agent orchestration, and customer-facing interfaces, with a strong emphasis on Python and platform engineering.

Role Overview

In this position, you will own the full path from idea to production while advancing agent execution quality and integrating AI workflows into the product experience. The work spans designing APIs, orchestrating agent runs, and building user interfaces that provide visibility and control.

Responsibilities

• Build and maintain the agent platform.
• Design APIs that power the product.
• Solve the hard parts of agent execution.
• Bring AI workflows to life in the product.
• Make the full system observable.
• Make quality measurable.
• Own the path from idea to production.

Required Qualifications

• Depth in Python and FastAPI, including shipped and operated production Python services.
• Ability to build well-designed FastAPI APIs and understanding of asynchronous programming, validation, authentication, and error handling.
• Strong systems fundamentals, including designing clear service contracts and data models.
• Comfort working with PostgreSQL and Redis.
• Thoughtful handling of concurrency, transactions, idempotency, and failure recovery.
• Experience turning AI into useful software, including reasoning about tool calling, context management, agent sessions, and boundaries between model behavior and application logic.
• Ability to deliver across the stack, including working in React, Next.js, and TypeScript and connecting interfaces to backend APIs.
• Care for product feel, including usability, accessibility, and performance.
• Experience making systems understandable using traces, logs, and metrics to investigate production issues and build instrumentation.
• Testing, security, and maintainability as part of shipping.
• Ownership and good judgment, including turning ambiguity into a practical plan and following through.
• Ability to explain tradeoffs and focus on solving customer problems.

Technologies

• Python, FastAPI
• React, Next.js, TypeScript
• Re state, Redis, PostgreSQL
• Datadog, OpenTelemetry (including tracing), structured logs, metrics, alerts
• Braintrust
• Asynchronous programming, authentication

Preferred and Ideal Traits

• Depth in Python and FastAPI.
• Strong systems fundamentals.
• Experience turning AI into useful software.
• Ability to deliver across the stack.
• A habit of making systems understandable.
• Ownership and good judgment.
• Plus: experience with Restate or Temporal, sandboxed agent execution, streaming systems, or AI evaluation tools.
• Healthcare or pharma experience is welcome but not required.
• You do not need experience with every tool in the stack.

Compensation

• Location: New York, NY (onsite)
• Base Salary Range: USD $200,000 - $300,000 per year
• Compensation Range: $200K - $300K (plus equity)

Benefits

• Medical, dental, and vision coverage
• Equity in an early-stage company
• Unlimited paid time off
• 401(k) with a 4% employer match
• Company-provided office lunches five days a week
• Stipend for professional development
• Relocation assistance
• Sponsorship for O-1, H-1B, and TN visas

What You Will Build

This role focuses on creating the foundation for how people work with AI, combining infrastructure that runs agents with product experiences that help customers direct, inspect, and use agent output. You will contribute to architecture and production tradeoffs, helping long-running agents operate reliably, streaming progress into responsive interfaces, and delivering the right context securely across every step of execution.

Collaboration and Scope

You will partner with product, design, Applied AI, and the broader engineering team to move work from initial idea through architecture, implementation, and production. The scope includes building new capabilities and improving the systems behind them so they are easier to operate and extend.