# Software Engineer
**Company:** [Reacher](https://hotfix.jobs/companies/reacher)
**Location:** San Francisco, CA
**Salary:** $125K-$175K
**Experience:** 3+ years
**Skills:** Python, FastAPI, Postgres, GCP, GKE, Ai Agent Frameworks, Llm Orchestration, Claude, Rest, GraphQL, Pytest, Playwright, Keda
**Posted:** 2025-11-13
> Build and own long-running AI agents automating multi-step workflows for TikTok Shop brands, alongside backend systems using Python, FastAPI, PostgreSQL, and GCP. Requires 3+ years backend experience with strong systems thinking for agent reliability.
## Job Description
## What You'll Do

### AI Agents
- Build and own long-running AI agents that automate complex, multi-step workflows for TikTok Shop brands.
- Design agentic systems end-to-end: orchestration, state management, error recovery, and observability.

### Backend & Infrastructure
- Own the backend systems your agents run on — design, implement, ship, maintain. **Python, FastAPI, PostgreSQL, GCP**.
- Contribute to our jobs infrastructure (**GKE-based, on-demand workers**) as needed to support agent reliability and scale.
- Set the bar for code quality and architecture decisions across the backend.

## You're a Fit If
- **3+ years** of backend engineering experience, ideally at a product-first company
- You've built automation, agents, or complex async systems — and you think deeply about failure modes
- Strong systems thinking, but applied toward agent reliability and workflow design, not just raw infra
- Comfortable with ambiguity and own problems end-to-end without hand-holding
- You've worked directly with customers or product teams and can translate vague requirements into technical decisions
- Resourceful: you figure things out, you don't wait for the answer to show up

## Technical Skills We Value

### AI and Agentic Development
- AI agent frameworks and LLM orchestration for building and managing long-running, multi-step agents
- **Claude** Code proficiency: context engineering (CLAUDE.md), custom commands, MCP integrations, subagent coordination
- Strong judgment for when to leverage AI vs. when to code directly
- Ability to decompose complex engineering work into tasks suitable for AI assistance
- Experience verifying AI-generated code for correctness, security, and maintainability

### Backend
- **Python, FastAPI, PostgreSQL**
- System design with a focus on async, long-running, stateful workflows
- **GCP**/cloud infrastructure (**GKE/KEDA**) — enough to own what your agents need
- API design (**REST/GraphQL**)

### General
- Testing frameworks (**Playwright, Pytest**) — especially for validating agent behavior end-to-end
- Code reviews, documentation, and establishing conventions

## Bonus Points
- You've been an early engineering hire at a startup
- You've built or maintained internal automation systems that ran at scale
- You've worked on AI agents, LLM pipelines, or autonomous systems
- You've shipped side projects or contributed to open source
- You're curious about e-commerce, creator marketing, or what AI will do to the agency business
**Apply:** https://hotfix.jobs/jobs/software-engineer-at-reacher-de2c6a5b-706b-47c6-9f2d-6d4deb5b7a7e
**Canonical:** https://hotfix.jobs/jobs/software-engineer-at-reacher-de2c6a5b-706b-47c6-9f2d-6d4deb5b7a7e