Job Description

At EX Squared LATAM, we are a digital
transformation company focused on building innovative, high-impact technology
solutions that drive real business value. We partner with top-tier clients
across global, regional, and local markets, fostering a collaborative,
inclusive, and innovation-driven culture where your ideas truly matter.

You use modern tools and frameworks (REACT and Node.js) and can code up an application from scratch and spin up a cloud infrastructure to host it.

You leverage modern tools and frameworks (REACT and Node.js) to architect and develop applications from the ground up, with a particular focus on enabling agentic workflows.

You design systems that empower autonomous agents to interact, make decisions, and execute tasks seamlessly building both the
application logic and the underlying cloud infrastructure needed to support scalable, dynamic agent operations.

You have a clear understanding of security and build it into any of your services.

Although you have full support of your teammates, you can participate in the research and solution to complex problems.

At the same time, you know when to ask for help and don’t let pride or ego get in the way, because after all, you are part of a team.

“Initiative” is your middle name, and you don’t give us a chance to chase you.

We are on the hunt for people who have an innovative mindset (you still dream of that moonshot project but are currently working on your side -gig)!

Is this you?

Location

• Fully remote
  position within LATAM.
  

Contract Duration

• Fixed
  term (6 months, could be extended).

Language Requirement

• Advanced English.

What you'll do

• Build and maintain automated evaluation pipelines (evals) for agent and skill outputs — including LLM-as-judge scoring, regression test suites, and golden dataset validation — so that prompt and model changes are measurable before they ship.
• Define strict input/output contracts for MCP tools and agent skills using typed schemas (TypeScript interfaces, JSON Schema); ensure tools handle edge cases, surface structured errors, and never return ambiguous output that an LLM must interpret.
• Own the prompt engineering lifecycle for assigned tools and skills — version-controlled prompt templates, parametric input injection, and structured system/user role separation — ensuring prompts are testable, reproducible, and free of implicit context drift.
• Contribute to the team's MCP tooling catalog — implement, test, and document MCP compatible API integrations (particularly GraphQL/Apollo and developer portal tooling); participate in rollout readiness reviews including SLO definition and support contact documentation.

What you'll bring

• 5+ years of working knowledge with modern frameworks and languages (REACT & NodeJS).
• Familiarity with OAuth2/OIDC, API key management, and access audit trail patterns.
• Experience with SSR applications.
• Experience with GraphQL and Apollo.
• Experience building or integrating MCP-compatible REST/GraphQL APIs; familiarity with tool interface contracts, versioning, and consumer documentation.
• Familiarity with SLO definition and basic observability practices (logs, metrics, distributed traces) for production API services.
• Hands-on experience with LLM structured output patterns: function/tool calling
• Familiarity with LLM evaluation approaches: building golden test sets, LLM-as-judge pipelines, and prompt regression testing; ability to quantify output quality with measurable metrics rather than vibes.
• Experience building agentic workflows — multi-step tool chains, state machine-based agents, or orchestration frameworks (LangGraph, LangChain, AutoGen, or custom) — with an emphasis on deterministic routing and graceful failure handling

What will make you stand out

• Demonstrated experience defining and rolling out engineering standards at a team or org level (coding conventions, PR workflows, testing mandates, API contracts).
• Experience managing and growing engineers — performance conversations, career development, structured feedback.
• Track record of driving AI tool adoption within an engineering team — not just using the tools but creating the onboarding, guidance, and culture around them.
• Comfort operating at the intersection of technical leadership and delivery management — you can write the ADR and run the retro.
• Familiarity with AWS & Docker.
• Experience with Styled Components.
• Experience with prompt caching, semantic routing, or output memorization strategies to reduce non-determinism at scale.
• Ability to instrument LLM calls with structured traces (input, output, latency, token count, tool calls invoked) using frameworks like LangSmith, OpenTelemetry, or custom logging; can define SLOs for agent task success rates.

Why EX Squared?

At EX Squared, we
believe in building strong, collaborative teams that create meaningful impact.
You’ll work alongside talented professionals across the region, contributing to
innovative projects while continuing to grow your career in a supportive, flexible,
and people-focused environment.

We value curiosity,
ownership, continuous learning, and open collaboration, creating an environment
where professionals can thrive, grow, and make a meaningful impact.

What We Offer

• Competitive compensation in
  USD.
• Access to learning platforms
  and professional development opportunities.
• Flexible work environment.
• Collaborative and
  multicultural teams.

Eligibility Note

This opportunity is
only available to candidates currently residing in Latin America or Brazil. Applications outside these regions will
not be considered.

Ready to become an EXpert?

Join our EXperts
team! #ApplyNow