At Northbound, we are building AI agents that automate complex workflows in global logistics and customs operations for large enterprise customers.

As Tech Lead - Agentic AI & AWS Backend, you will be the driving force behind how our AI agents reason, plan, and act. You will own the design and evolution of our agentic architecture – the orchestration layer that coordinates multi-step workflows across regulatory lookups, document validation, downstream system integrations, and complex decision-making.

You will work directly with the technical founder to translate messy, real-world logistics problems into reliable, production-grade agentic systems. This means designing the right abstractions – deciding when to use single vs. multi-agent architectures, how to manage context and memory across long-running workflows, and how to build guardrails that keep autonomous systems safe and predictable.

This is a hands-on role for someone who lives and breathes the rapidly evolving world of agentic AI and wants to apply it to hard, real-world problems at enterprise scale.

What You’ll Do:

• Design and build the agentic orchestration layer that powers our core product – coordinating planning, tool use, memory, retrieval, and self-verification across multi-step workflows.
• Architect multi-agent systems with clear abstractions – deciding when to decompose into specialized agents vs. single-agent flows, and how agents communicate and hand off tasks.
• Design and build tools that agents use to interact with external systems – APIs, databases, document processors, and downstream services.
• Design guardrails, failure handling, and human-in-the-loop mechanisms for autonomous workflows.
• Integrate and operate LLMs in production, making informed decisions on model selection, cost-performance trade-offs, and multi-model orchestration strategies.
• Build and maintain document processing pipelines – extracting structured data from PDFs, Excel files, and scanned documents as inputs to agent workflows.
• Design and operate resilient backend systems on AWS, including serverless architectures, to support agent workflows at scale.
• Contribute to the Python backend platform, ensure software quality through reviews, testing, and CI/CD, and mentor engineers across the team.

Required Skills & Experience:

Agentic AI (Primary)

• At least 1 year of professional experience building and deploying LLM-powered agentic systems in production environments.
• Hands-on experience with LLM platforms such as AWS Bedrock, OpenAI, and Anthropic Claude.
• Strong understanding of orchestration patterns: multi-agent coordination, tool use, planning, memory management, context routing, and state machines.
• Experience designing and building tools for agents – writing reliable interfaces that let agents interact with APIs, databases, and external services.
• Experience with the LangChain ecosystem – LangGraph, LangSmith, and observability tools like LangFuse. Familiarity with alternatives like CrewAI, AutoGen, or OpenAI Agents SDK is a plus.
• Experience implementing structured evaluation frameworks for measuring agent reliability, accuracy, and cost.
• Working knowledge of reflection, ReAct, and self-correction patterns.
• Familiarity with Model Context Protocol (MCP) for standardized tool integration and agent-to-system communication.

Engineering Fundamentals

• 7+ years of professional software development experience with strong Python and backend engineering skills.
• Proven experience shipping and operating production systems in real-world environments.
• Solid understanding of API design, distributed systems, and event-driven architectures.
• Experience with document processing pipelines – PDF/Excel extraction, OCR, and familiarity with services like AWS Textract or similar alternatives.

Highly Valued

• Public GitHub projects or open-source contributions related to agentic AI, LLM tooling, or agent orchestration.
• AWS experience, especially serverless services (Lambda, API Gateway, S3, DynamoDB, SQS). AWS certification is a strong plus.
• Experience with Flask/FastAPI and RESTful API design.
• Knowledge of Docker and containerized deployments.
• Infrastructure as Code (AWS SAM, Terraform, CloudFormation).
• Experience working in early-stage or high-growth startups.
• Exposure to enterprise security and compliance environments.

What We Offer:

• Direct collaboration with the Technical Founder on core product and AI strategy.
• Real ownership of the agentic AI platform powering global enterprise customers.
• Opportunity to shape the technical direction of an enterprise AI company at an early, high-impact stage.
• A modern, well-equipped office in Berlin Kreuzberg, shared with a strong tech community.
• A focused, high-trust engineering culture with minimal bureaucracy.
• Competitive compensation and ESOP/VESOP package.