About the Role and Mission
You will be part of the team responsible for Seven, Inter's virtual assistant, and will act directly in building our Generative AI agent platform. Your role will be central to Inter's AI First transformation – developing scalable solutions with LLMs, orchestrating intelligent agents, and ensuring that artificial intelligence truly improves our customers' lives. If you want to be at the forefront of innovation in one of Brazil's largest digital banks, this is your opportunity.
Responsibilities
- Develop and evolve Inter's Generative AI agent platform, ensuring scalability and resilience.
- Build and orchestrate intelligent agents using LangChain, LangGraph, and models from OpenAI and Anthropic.
- Implement RAG pipelines with pgvector to enrich agent responses with relevant context.
- Develop and maintain APIs with FastAPI and integrations via REST, connecting agents to bank systems.
- Contribute to the evolution of Seven, Inter's virtual assistant present in the app for millions of customers.
- Monitor system health and performance with Grafana and New Relic, acting proactively to identify problems.
- Work on reducing inference costs, evaluating models, caching strategies, and prompt optimization.
- Collaborate with multidisciplinary teams in developing solutions that put Inter at the forefront of AI First.
- Participate in Scrumban agile rituals, contributing with estimates, reviews, and retrospectives.
Requirements
Mandatory:
- Solid experience with software development in Python, Java, or .NET.
- Experience with REST API architecture and development.
- Knowledge of containerization and orchestration with Docker and Kubernetes.
- Familiarity with relational and non-relational databases (MongoDB, Redis, PostgreSQL).
- Experience with agile methodologies (Scrum, Kanban, or Scrumban).
- Ability to learn and quickly adapt to new technologies, especially in the context of Generative AI.
Differentiators:
- Experience with development or orchestration of AI agents using LangChain or LangGraph.
- Knowledge of RAG and vector databases like pgvector.
- Familiarity with LLM models from OpenAI and Anthropic and good LLMOps practices.
- Experience with FastAPI for high-performance API development.
- Knowledge of observability tools like Grafana and New Relic.
- Experience with cloud platforms, especially AWS Bedrock.
- Familiarity with AI-assisted development tools like Cursor and Claude Code.
Desirable:
- Intermediate English for reading technical documentation and following updates in the AI ecosystem.
- Experience with mobile development.
- Experience in technology companies, fintechs, or digital banks.
- Knowledge of LiteLLM for managing and routing multiple LLM models.
- Familiarity with the MCP (Model Context Protocol).
- Experience with Guardrails for validation and security of LLM outputs.
- AWS Certifications, especially related to machine learning or Bedrock.