Software Engineer - Applied AI

🇧🇷 São Paulo / 🇲🇽 Mexico City / 🌎Remote5+ years

A little bit about us:

We are Belvo, an open finance API platform with the bold vision of democratizing access to financial services in Latin America. We enable any financial innovator to access and interpret financial data, as well as initiate payments from their end-users accounts. We’re turning the messy complexities of the Latin American financial ecosystem into a modern set of tools to access and interpret data and move money in a seamless and secure way.

We’re a highly-technical, passionate, and driven team. We are more than 90 people and our team currently represents 20 nationalities. We have offices in São Paulo and Mexico City – while a large portion of us work remotely.

We are tackling a very stimulating problem: connecting fintech innovators with legacy financial infrastructure. We strive to go beyond the limits of what is possible today and to do so in an elegant and developer-first way.

Since starting our adventure in May 2019, we have raised $71m from the leading VC investors globally.

You can read more about our company here and about our team and culture here. Also, head to our blog for more news about what we’re building and how we work.

About the team:

We work in cross-functional, autonomous teams. We follow continuous delivery best practices executed on top of a modern technology stack.
Our products are built for developers, by developers. Technological excellence is at the heart of what we do.
We are pragmatic and customer-focused. We strive to find the right set of trade-offs in order to validate our hypothesis as early as possible, iterating on our products based on customer feedback.
We communicate transparently. We do weekly all-hands where we get together to discuss company performance and goals.
While we are global and remote-friendly, we also operate from our vibrant offices in CDMX and Sao Paulo. To accommodate the various time zones in which we are based, we ensure we’re always synced up for at least 3 hours per day.
Also, we are backed by some of the leading investors in Silicon Valley and Latin America, including Founders Fund, Kaszek Ventures, and YCombinator.

The challenge

The Intelligence Squad is Belvo's AI enabler team. We provide the platforms, tools, and patterns that allow every team at Belvo to build their own AI workflows. We also build and ship new product bets rapidly within the fintech space. One week we might be shaping a new internal assistant to help our teams ship faster; the next, we could be launching a customer-facing AI experience from scratch and validating it with real users. We move fast, ship MVPs, learn what works, and double down—all while strengthening shared foundations (agent abstractions, observability, evaluation tooling, guardrails) so the next bet is even faster to build.

In this role, you will be hands-on across the full lifecycle of AI products: prototyping, evaluating, deploying, and operating them in production alongside the squad. You will work AI-first, actively integrating agents and copilots into your daily workflow rather than relying on them just as chat assistants. This is an opportunity where building with AI and building with the help of AI are both core to the day-to-day work.

Responsibilities

Product Delivery: Contribute hands-on across the full lifecycle of AI products, from prototyping new initiatives to deploying and maintaining production AI services for internal teams and external customers.
Agent Lifecycle Management: Plan, implement, and maintain end-to-end agents. This includes selecting the right models, designing prompts, setting up Model Context Protocols (MCPs) and tool integrations, running evaluations, and ensuring production health.
AI-Native Workflow: Operate AI-natively every day. Actively utilize AI agents (e.g., Claude Code, Cursor, Codex, custom pipelines) to plan, implement, review, and debug your work.
Platform Foundation: Contribute to the foundational platform other Belvo squads use to ship AI, focusing on agent abstractions, conversation and state storage, durable workflow orchestration, streaming interfaces, observability, and guardrails.
Technical Exploration: Stay at the forefront of the industry by testing new models, agentic frameworks, and MCP-native patterns, selecting options based on pragmatic production reliability, cost, and latency.
Observability & Monitoring: Instrument and monitor AI pipelines using Langfuse and Datadog to debug failed agent runs, detect regressions, and maintain production health.
Team Enablement: Support enablement efforts that help other Belvo squads adopt AI safely through reference implementations, integration patterns, and shared tooling.

This position may be for you if:

5+ years of experience in software engineering or building technical products.
You have contributed to building and maintaining LLM-powered systems (agentic systems) in production settings.
You are comfortable working with agentic frameworks (e.g., Pydantic AI, LangGraph, Claude Agent SDK, AI SDK) and applying them effectively in context.
You are familiar with the Model Context Protocol (MCP) and its core design considerations.
You actively use coding agents (e.g., Claude Code, Cursor, Codex, GitHub Copilot) in your daily workflow and consistently look for ways to amplify your output with them.
You are comfortable reading agent traces, debugging AI pipelines, and utilizing AI observability tooling (e.g., Langfuse, Datadog).
You bring strong engineering skills in Python (ideally with FastAPI or Django) and focus on writing maintainable, well-tested code.
You have experience with cloud infrastructure (ideally AWS) and infrastructure-as-code (e.g., Terraform), and feel comfortable deploying and maintaining your own services.
You are curious about new models and tools, yet thoughtful about what is practical for production use.
You value sharing progress early, collaborating closely with stakeholders, and viewing learning as an iterative process.
You thrive in fast-paced or small-team settings where ownership, collaboration, and adaptability are critical.
You are proficient in written and spoken English.

Amazing if:

You have previous experience in Fintech or SaaS product environments.
You have built custom AI-assisted development workflows, orchestrating multiple agents to plan, build, review, or debug your own work.
You have experience designing thoughtful evaluation approaches for AI features (e.g., golden datasets, LLM-as-judge, regression testing).
You have experience building voice or multimodal AI experiences (e.g., STT/TTS, realtime APIs).
You are comfortable working with frontend code (TypeScript/React) when needed to ship a full-stack prototype (frontend specialization is not required, but it is a plus).

Our tech stack:

The AI stack changes every day, so flexibility is essential: learn fast, iterate quickly. We currently rely on Pydantic AI for most of our agents and continuously evaluate new frameworks like Claude Agent SDK, dspy, and others.
We primarily use modern Python on the backend (FastAPI/Django).
We use TypeScript and React on the frontend.
We run our infrastructure on AWS, favoring managed services wherever possible so we can focus on business problems.
We observe and monitor AI pipelines with Langfuse and our broader services with Datadog.

Our process steps:

At Belvo every hire is so important to us, that we share the decision to hire as a team.