#9 AI Backend Engineer Senior
Summary
Builds the backend of a large organization's internal AI platform that other developer teams use to create AI agents: Python/FastAPI services, cross-team SDKs, MCP servers, a unified LiteLLM inference layer, and RAG pipelines, plus supporting developer adoption. Requires 5+ years backend and 2+ years agentic AI experience.
🚀 Join Our Remote Data Products & Machine Learning Startup! 🚀
At Muttdata, we build innovative Data Products and Machine Learning solutions that help companies solve complex business challenges. As a fast-growing, remote-first startup, we're passionate about technology, collaboration, and continuous learning.
This team builds the platform other teams use to create AI agents: SDKs, backend services, MCP servers, and a unified inference layer. Your users are other developers, so part of the job is supporting them through adoption.
🚀 What We Do
- Leveraging our expertise, we build modern Machine Learning systems for demand planning and budget forecasting.
- Developing scalable data infrastructures, we enhance high-level decision-making, tailored to each client.
- Offering comprehensive Data Engineering and custom AI solutions, we optimize cloud-based systems.
- Using Generative AI, we help e-commerce platforms and retailers create higher-quality ads, faster.
- Building deep learning models, we enhance visual recognition and automation for various industries, improving product categorization, quality control, and information retrieval.
- Developing recommendation models, we personalize user experiences in e-commerce, streaming, and digital platforms, driving engagement and conversions.
🌟 Our Partnerships
- Amazon Web Services
- Astronomer
- Databricks
🌟 Our Values
- 📊 We are Data Nerds
- 🤗 We are Open Team Players
- 🚀 We Take Ownership
- 🌟 We Have a Positive Mindset
Responsibilities 🤓
- Design and develop backend services and agents using LangChain and LangGraph.
- Manage the agent lifecycle, from development and deployment to observability, monitoring, and continuous improvement.
- Build and maintain cross-team SDKs.
- Develop, publish, and secure MCP servers, including their authentication.
- Unify agent inference through LiteLLM or a similar proxy.
- Implement guardrails, PII masking, and Agent Skills.
- Build and optimize RAG pipelines, vector databases, and context management.
- Support the teams that consume the platform, helping them adopt it and unblocking them along the way.
Required Skills 💻
- 5+ years of backend development experience and 2+ years working with agentic AI.
- Strong command of Python and FastAPI (or similar frameworks).
- Experience with LangChain, LangGraph, and modern agent-building stacks.
- Solid knowledge of MCP across its full lifecycle (development, publishing, and authentication).
- Experience with LiteLLM or similar proxies.
- Experience with vector databases, RAG, and context optimization.
- Knowledge of guardrails, PII masking, and Agent Skills.
- Experience building and maintaining SDKs.
- Ease supporting other developers.
- Stack: Python, FastAPI, LangChain, LangGraph, Langfuse, MCP, LiteLLM.
Nice to have 💻
- TypeScript.
- Kubernetes.
- AWS.
🎁 Perks
- 🌍 Remote-first culture – work from anywhere!
- 🚀 In-Company English Lessons.
- 💪 Wellhub or sports club stipend to stay active
- 🚀 AWS, DBT, Google Cloud, Azure & Databricks certifications fully covered
- 🍕 Food credits via Pedidos Ya – because great work deserves great food.
- 🎂 Birthday off + an extra vacation week (Mutt Week! 🏖️)
- 🤝 Referral bonuses – help us grow the team & get rewarded!
- ✈️🏝️ Annual Mutters' Trip – an unforgettable getaway with the team!
- 👶 Monthly Childcare Reimbursement – Because supporting families matters too
Skills
As published by lever · 10 questions · 8 written answers
Basics
Resume/CV, Full name, Email, Phone, Current location, Current company, LinkedIn URL, GitHub URL, Other URL
Pick from a list (2)
- Where did you hear about us? optional
- English Level
Written answers (8)
- Tell us about something you built for other developers to use, such as a library, an SDK, an agent template, or an MCP server. How many teams used it, and how did you support them?
- Tell us about an agent you took to production. What did you build it with, and how did you find out when it answered badly?
- Have you built your own MCP server? Who connected to it, and how did you control who could use each tool?
- Which models were you working with, and what had to change to move from one provider to another?
- How did you prevent an agent from exposing customers' personal data?
- Study/Degree optional
- Gross salary expectations (ARS)
- The position requires coming into the office once a week; do you live in Buenos Aires? This is a mandatory requirement for the role