AI/ML Engineer

About the role

binbash is an AWS Advanced Tier Services Partner. We build the DevOps, data and AI platforms that startups run their companies on, and our AI/ML practice ships the GenAI half of that — retrieval systems, agents, and the evaluation that decides whether either is fit to put in front of a user.

You would join that practice as a senior individual contributor: hands on the systems, working alongside the pre-sales architects who scope them and the delivery engineers who run them. Not a research seat — the work is production GenAI for clients whose users notice when it is wrong.

Requirements / Technical Skills

• Strong Python, and comfort reading and writing infrastructure as code (Terraform)
• Production experience with a hosted model API — Amazon Bedrock ideally, another provider otherwise
• Retrieval systems end to end: chunking, embeddings, a vector store (OpenSearch, pgvector or similar), and the reranking that makes results usable
• Agent and tool-calling architectures, orchestrated with AWS Step Functions, Amazon Bedrock AgentCore or an equivalent
• Model evaluation you can defend: task-level test sets, regression suites, and the judgement to say when a system is not ready
• Amazon SageMaker for training, tuning or hosting where a managed model is the wrong answer
• Excellent verbal and written communication skills (English)
• Enthusiasm for working in a startup environment and the ability to be cross-functional

Key Responsibilities

• Design and ship GenAI systems on AWS — retrieval, agents, and the pipelines that feed them
• Build the evaluation before the feature, so a change to a prompt, a model or a corpus has a number attached to it
• Advise clients on model selection, context strategy and the cost of both, with the same candour we would want
• Take a system from a working prototype to something the delivery team can operate, monitor and hand over
• Feed what you learn back into how we scope and price GenAI work

Where you are based

Fully remote, as we have been since binbash was founded. We would prefer Argentina, Uruguay or Colombia, because that is where most of the team already sits and it keeps us in one working day — but it is a preference and not a filter: make the case for where you are.

Availability to travel is a strong plus rather than a requirement — team gatherings, and the AWS events we run and attend.

What we Offer

• Opportunity to work with a leading AWS Advanced Tier Services Partner.
• A collaborative and innovative work environment.
• Continuous learning and professional development opportunities.
• Competitive compensation in USD.
• AWS certifications plan and reimbursements.
• Remote, with 30–40 hour weeks, no weekends, no on-call rotation and no work on your national holidays.