SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Labra builds cloud-native B2B SaaS products powered by AI agents that help enterprises accelerate cloud transformation. This is a hands-on Senior Software Engineer role with real scope across three interconnected areas: AI agents and LLM features, backend sync engine, and React-based product development.
You will own high-impact systems end-to-end—from half-formed customer problems through production deployment and on-call support. Your depth can be in one area (AI agents, backend, or frontend), but you must be comfortable working across all three. You'll join a pod that balances feature velocity with stability, taking problems from conception to reliable production systems.
Key Responsibilities:
AI Agents & LLM Features: Build production AI agents with RAG, tool calling, and MCP-based workflows. Design context management (memory, compression, retrieval, prompt optimization). Optimize for time-to-first-token (TTFT), latency, token usage, and cost through caching, parallel execution, and LLM call reduction. Ship streaming AI experiences over SSE and WebSockets.
Sync Engine & Integrations: Design and ship modules for business system integrations (CRM, ERP). Make high-volume, bidirectional sync between CRMs and cloud partner programs fast, correct, and recoverable. Build event-driven processing on AWS SQS, EventBridge, SNS, Lambda, and Step Functions with robust retries, idempotency, and dead-letter handling.
Full-Stack Product: Build customer-facing React/TypeScript features backed by well-designed REST APIs in Python or Go. Implement authentication, authorization, RBAC, and secure third-party integrations. Profile and optimize slow endpoints and queries; load-test before customers hit limits. Write automation and unit tests.
Production & Engineering Quality: Add metrics, logs, traces, and alerts; cut alert noise. Join on-call rotation, lead incident debugging, ship root-cause fixes. Improve scalability, reliability, and cloud cost efficiency on AWS. Write design docs, review code, mentor junior engineers.
Tech Stack: Python (primary), TypeScript, Go; React frontend; FastAPI/Flask/NestJS backends; Amazon Bedrock, Claude, LangChain/LangGraph; PostgreSQL, MongoDB, DynamoDB, OpenSearch; AWS (ECS Fargate, Lambda, SQS, SNS, EventBridge), GCP, Azure; Docker, Terraform, CloudFormation.
Culture: Fast-paced startup shipping to large enterprises. You act as an owner, close the loop from merge to production confirmation, move fast without breaking trust, handle ambiguity, raise problems early, keep learning, and lift the team.
Requirements:
- 5+ years professional software engineering (ideally 5+ in B2B SaaS production)
- Strong Python or Go; TypeScript with willingness to work primarily in Python
- Hands-on production React and TypeScript experience
- Solid REST API design; experience with microservices, distributed systems, asynchronous processing (ideally AWS SQS, SNS, Lambda)
- Depth in at least one relational database (PostgreSQL or MySQL) and one NoSQL store (MongoDB or DynamoDB): schema design, indexing, query tuning
- Experience building or integrating LLM features, AI agents, RAG, or tool-calling systems; working grasp of context windows, token management, retrieval
- Understanding of TTFT, streaming, latency optimization, caching
- Hands-on AWS (and Azure or GCP) production deployments; Docker, Git, CI/CD, automated testing
- Comfort debugging production issues from logs, metrics, traces
- Track record owning a feature or service from design through production support
- Clear written communication: design docs, tickets, updates
- Proven track record mentoring and developing junior team members
Nice to Have:
- MCP servers/clients, Amazon Bedrock, Bedrock AgentCore
- Agent orchestration with LangChain/LangGraph; LLM evaluation, guardrails, agent observability
- Workflow orchestration (Temporal, Airflow, Step Functions, Celery)
- CRM integration experience (Salesforce, HubSpot APIs)
- OpenSearch or vector/search engines for retrieval
- Infrastructure as code (Terraform, CloudFormation)
- Load testing (k6, Locust) and observability tooling
- Cloud partner programs or marketplaces (AWS ACE, Azure, GCP)
