Software Development Engineer -3 (SDE-3)

About Loop: Loop AI is an agentic AI platform for the back office of the real economy. Today we build end-to-end agentic workflows for enterprises; our goal is for users to build and manage agents themselves. Founded in 2022, we're trusted by 300+ brands powering thousands of locations across the US, backed by a $14M Series A led by Nyca Partners, with teams in New York, San Francisco, Tampa, and Bengaluru.

We're built by a world-class team. Several of us are ex-founders who scaled our own companies to multiple millions in ARR before joining hands at Loop to execute on a bigger shared vision — alongside operators and AI engineers from Stanford, Harvard, Wharton, Uber, Google, Amazon, and McKinsey.

We believe agentic products are the future of the software–human interface, and restaurants were only the beginning. Flat hierarchy, lean high-impact team, radical ownership, no bureaucracy — this is your chance to help steer the rocketship.

About the Role: As an SDE-3, you set the technical direction for a problem area and take it to production. Our stack is agent-native. You'll work across product surfaces, data and BI systems, and the agentic infrastructure behind them - long-running workflows that have to be repeatable and idempotent, orchestration and tool-calling, and the distributed services that feed them. Agents are a first-class part of the stack, not the only thing you'll touch.

You work directly with product and cross-departmental leads - GTM, CS, Sales & Leadership - to scope product solutions and get them through production.

We hire for end-to-end ownership, product curiosity, and the hunger to raise the bar for everyone around you.

Key Responsibilities:

-
Lead system design: Own the architecture for major features and services. Write the design down, drive it through review, and make the tradeoffs on reliability, latency, and cost explicit before anyone writes code.

-
Raise the engineering bar: Run thorough code reviews, set standards for testing and observability, and mentor SDE-1s and SDE-2s into stronger engineers.

-
Build products end to end: Take ambiguous problems from conversation to production. Scope them, break them into shippable slices, and ship the hardest parts yourself. Communicate the present the same to cross-departmental leads.

-
Implement AI agent workflows: Build and integrate LLM-powered agentic features: tool/function calling, prompt orchestration, retrieval, and the guardrails that keep autonomous actions safe and auditable.

-
Feed the agents: Develop & manage ingestion streams, ETL pipelines, and the data architectures that give agents the data they need to automate workflows, and execute operational tasks.

What We're Seeking:

-
6+ years of full-time experience building and shipping production backend systems and web applications, including systems you designed and owned end to end.

-
Deep fundamentals in data structures, algorithms, and distributed system architecture. You can design a system, defend the tradeoffs, and spot the failure mode nobody else raised.

-
A track record of leading design reviews and code reviews, and of making the engineers around you better.

-
Real experience building data pipelines, ideally agentic ones that feed and are driven by LLM workflows.

-
Hands-on with agentic application development: model APIs, prompt design, tool-calling patterns, and evaluation. You also code with AI tools daily without dropping the bar on correctness and review.

-
Comfort with cloud services (GCP preferred) and event-driven architectures. Exposure to BI or analytics platforms is a plus.

-
Product-first thinking and 0-to-1 experience. You work directly with product folks across functions and geographies, decide what is worth building, and debug and optimize what you own without waiting to be asked.

-
Experience working directly with customers - hearing their issues first-hand and using that to shape what you build.