Data Science Specialist – Generative AI, Agents

This is a fully remote position, open to applicants in Brazil.

📋 Description

• Conduct research and experimentation on Generative AI methodologies, encompassing agents, LLMs, SLMs, multimodal models, RAG, and fine-tuning techniques.

• Collaborate with various business sectors within the bank to identify challenges, formulate hypotheses, and create AI solutions addressing real-world issues.

• Oversee hands-on projects from the experimentation phase to implementation and validation, integrating models, data, APIs, tools, memory, retrieval systems, and safety mechanisms.

• Design experiments, datasets, baselines, metrics, and evaluation standards, comparing internal solutions with those from partners and market competitors.

• Assess agents comprehensively and analyze failures, taking into account quality, grounding, memory, tools, robustness, cost, and performance metrics.

• Create reusable components, benchmarks, methodologies, and accelerators.

• Contribute to strategic projects concerning agent memory, evaluation, and monitoring processes.

• Embed requirements for quality, security, privacy, and Responsible AI practices.

• Share insights and act as a technical resource for the Center of Excellence (CoE) and various business sectors.

• Develop, assess, and improve Generative AI and agent solutions for banco BV’s GenAI CoE.

• Engage with internal teams, partners, corporations, universities, and researchers.

⛳️ Requirements

• Extensive experience in Data Science, Generative AI, and the creation of agents or similar AI solutions.

• Proficient in Python, software development best practices, testing methodologies, and API integration.

• Background in experimentation, metrics, model evaluation, prompts and agents, results analysis, and error investigation.

• Familiarity with RAG, embeddings, vector databases, chunking, retrieval techniques, reranking, and grounding concepts.

• Experience with agent frameworks such as ADK or comparable technologies.

• Understanding of planning, memory, tools, orchestration, and multi-agent workflows.

• Knowledge of LLMOps and observability, including tracing, versioning, and monitoring quality, cost, and latency.

• Experience with cloud platforms and AI services, preferably GCP and Vertex AI, or equivalent platforms.

• Ability to monitor technical and scientific advancements and convert them into practical solutions.

• Scientific mindset with the capacity to turn challenges into testable hypotheses.

• Curiosity towards applied research and emerging technologies.

• Technical expertise, a hands-on approach, and independence when tackling loosely structured problems.

• Skill in comparing options and making decisions based on evidence.

• Critical view on applicability, quality, limitations, and associated risks.

• Effective communication and collaboration with technical professionals, executives, business teams, and external partners.

• Experience with fine-tuning, PEFT, LoRA, quantization, or open-source models.

🏝️ Benefits

• Commitment to the inclusion of Black individuals, women, professionals with disabilities, LGBTQIA+ individuals, and people of all ages.

• Recognized as the Best Financial Institution to Work For in GPTW’s 2025 ranking.