Senior AI Engineer

Sr. AI Engineer

Job Description

Join Us to Shape the Digital World. We’re passionate about delivering cutting-edge technology to some of the world’s top startups and companies, powered by diverse and empowered teams of technologists eager to drive change.

As a fairness-driven organization, we are committed to creating a safe and inclusive environment where everyone, regardless of background, is treated with respect and equity.

We value people with strong technical skills who are collaborative, curious, results-driven, and take ownership. We embrace people who want to be themselves, enjoy daily flexibility, and are eager to grow, learn, and make a difference wherever the opportunity arises.

If this resonates with you, we encourage you to apply for the role of Sr. AI Engineer. We’re seeking exceptional talent to work on immersive client projects that will challenge and hone your skills.

* Advanced English language proficiency is required for the role.

About The Role:
We are looking for a Senior AI / Backend Product Engineer to join our team and help build and scale a conversational AI platform for business workflows. You will work on AI systems that understand context, retrieve information from internal services, use tools to complete multi-step tasks, and support real-time product experiences.

This is a hands-on role at the intersection of backend engineering, agentic AI, product integration, and platform reliability — working closely with product managers, designers, data scientists, and engineers. The stack is Python, LangGraph, and AWS. The team is senior, distributed across Europe and LATAM, and treats engineers as peers.

Your Responsibilities:

• Design and implement AI agentic architectures tailored to customer engagement and automation use cases, including routing, planning, tool calling, memory, and domain-specific workflows.

• Build and maintain asynchronous backend services and real-time streaming interactions (SSE, WebSockets) between AI systems and product interfaces.

• Develop and maintain pipelines for agent reasoning, memory, planning, and tool use — integrating LLMs, MCP servers, RAG systems, and external APIs through secure, typed, and resilient service contracts.

• Implement reliability patterns for production AI: timeouts, retries with exponential backoff, circuit breakers, graceful degradation, caching, rate limiting, and concurrency management.

• Evaluate models and agents to improve performance, reliability, and alignment — including golden datasets, regression checks, LLM-as-judge scoring, feature flags, and guardrail pipelines.

• Implement guardrails for permissions, sensitive actions, unsafe inputs, hallucination risk, prompt-injection mitigation, and tool misuse.

• Standardize agent observability with structured traces (LangSmith, Langfuse, OpenTelemetry), turning opaque LLM failures into reproducible incidents with clear rollback paths.

• Collaborate with product, frontend, platform, and data teams to turn AI capabilities into useful and safe user experiences.

• Stay ahead of emerging trends in agent-based AI — including MCP, AG-UI, and agentic development workflows — and proactively bring innovative ideas into the product.

Technical Skills (Must-haves):

• 8+ years of software engineering experience, with a meaningful backend systems foundation (APIs, databases, distributed systems) and a focused shift to AI/agent development in the last 1–2 years.

• Strong production Python skills: async/await, asyncio, FastAPI, Pydantic, dependency injection, typed codebases (mypy, Ruff), and pytest including pytest-asyncio.

• Hands-on experience building LLM applications, AI agents, and RAG systems in production — not just PoCs. Must include real ownership: you built it, operated it, and fixed it when it broke.

• Solid experience with agent orchestration frameworks, specifically LangChain and LangGraph (state machines, conditional routing, tool nodes, checkpointing).

• RAG implementation depth: hybrid retrieval (BM25 + dense embeddings), reranking, chunking strategies, vector databases (pgvector, FAISS, Weaviate, Qdrant), and retrieval failure debugging.

• Familiarity with MCP (Model Context Protocol): understanding of tool contracts, server design, and agent-to-tool integration patterns.

• Reliability engineering applied to LLM/agent systems: retry logic, circuit breakers, fallback paths, streaming failure handling, and cost/token optimization.

• Agent evaluation and observability: golden datasets, regression testing, LLM-as-judge, tracing with LangSmith or Langfuse, structured logging, and alerting.

• Cloud environment experience (AWS preferred): ECS, Lambda, SQS, RDS/PostgreSQL, S3, Docker, Kubernetes.

• Strong communication and teamwork abilities in a fast-paced, distributed, collaborative environment.

Nice to have:

• Experience with MCP server development, OAuth-based MCP authentication, or MCP gateway architecture.

• Familiarity with real-time streaming patterns: SSE, WebSockets, AG-UI or equivalent agent-to-UI protocols.

• Background in human-in-the-loop systems, AI safety, or alignment practices.

• Active use of AI coding tools (Claude Code, Cursor, Codex) and specification-driven development (SDD) methodologies.

• Contributions to open-source agent frameworks or NLP libraries.

• Experience with reinforcement learning or agent simulation environments.

A happy team makes all the difference. That’s why we offer:

• Payment in USD

• A truly flexible work schedule

• A Non-Working Pay Days Policy

• Learning Budget

• An opportunity for you to help create change in the industry

• And more!
  
  

Athenaworks is an inclusive and safe organization that values your technical skills, work experience, collaboration abilities, and potential to grow in your career. We proudly welcome self-taught individuals, as we focus solely on your ability to deliver exceptional work. We will never consider any irrelevant personal or professional aspects of your life.