Senior/Staff Software Engineer

Sandbar•New York, NY

4h•$190,000 - $260,000

About The Position

Sandbar is seeking a generalist engineer to connect its Stream product with the software people already use. Stream is a private voice ring that listens, remembers, and helps users think through conversation. The engineer will determine reliable ways to connect applications to Stream and make those connections dependable for everyday use. The role owns integration work end to end, works directly with the founders, contributes to Stream’s real-time Python services, and helps shape how others build on Stream.

Requirements

At least 5 years of software development experience.
Strong Python skills and comfort owning a live, real-time service.
Production experience shipping third-party integrations, including OAuth 2/OIDC, token refresh and revocation, webhooks, API versioning, and vendor quirks.
Understanding of software architecture sufficient to work with applications the candidate did not write.
Experience integrating tools into an LLM agent used by real users and accountability for actions completing.
Hands-on MCP experience.
Generalist range across backend, authentication, client code, and agent logic; ability to take an open-ended problem through to a shipped result without a spec.
Product sense for software that acts on someone’s behalf, including when the user is not looking at a screen.
Precise communication and ability to make tradeoffs explicit.

Nice To Haves

Connector or plugin work at an AI product.
Browser automation experience, such as Playwright or Chrome DevTools Protocol.
Integration security experience, including scoped credentials, secret storage, prompt-injection defenses, and audit logging.
Voice or multimodal assistant experience.
Developer documentation that other engineers have built on.

Responsibilities

Build integrations between Stream and the apps, services, and AI tools people already use.
Design how tools plug into Stream’s agent, including tool definitions, result shaping, confirmation, and context passing.
Own the integration layer end to end, including OAuth and token lifecycle, least-privilege scopes, webhooks, rate limits, retries, and real-world failure modes.
Build and operate MCP servers and clients in production.
Develop reliable integrations for software without clean APIs.
Instrument integrations so breakage is caught before users notice.
Contribute to the core backend and participate in its on-call rotation.