We are looking for a **Senior Data Scientist** to work on the development and evolution of **applied AI solutions**, with a focus on **AI agents** (e.g., RAG, tool-use, memory, evaluation, and observability). You will work with data from multiple sources and collaborate with engineering and product teams to transform business needs into measurable, reliable, and scalable systems—with attention to **quality, cost, and performance**. **Responsibilities and duties** * Participate in defining and validating data- and experiment-driven business hypotheses; * Design and execute advanced analyses, metrics, and evaluation strategies for agents (e.g., golden sets, regression tests, failure analysis); * Build and maintain data pipelines and training/testing datasets, focusing on reproducibility and quality; Develop and fine-tune ML models when needed (baseline → production), and integrate them with generative AI solutions; * Support the design and improvement of RAG: chunking strategy, metadata, deduplication, ranking signals, knowledge base updates; * Implement and monitor observability: structured logs, quality metrics, dashboards, and alerts; * Collaborate with engineering and product teams on architecture definition, requirements, trade-offs, and roadmap planning; * Review PRs, share knowledge, and mentor less experienced colleagues. **Requirements and qualifications** * Advanced proficiency in Python and SQL (CTEs, window functions, basic optimization); * Solid foundation in applied statistics and evaluation (metrics, biases, testing, interpretation); * Practical experience in Machine Learning (training and validation, feature selection, metrics, deployment, and integration); * Experience with engineering best practices: versioning, testing, documentation, and observability; * Ability to structure problems and make decisions involving trade-offs (quality vs. latency vs. cost); * Experience working in cloud environments (AWS/GCP/Azure) and/or production pipelines. **Preferred qualifications** * Experience with **generative AI**, RAG, and agents (prompting, tool-use, memory, guardrails); * Experience with embeddings and vector search mechanisms (FAISS, pgvector, Milvus, Pinecone); * MLOps: experiment tracking (MLflow/W&B), CI/CD for models, drift monitoring; * Building evaluation datasets and labeling (process, guidelines, agreement); * Experience with A/B testing and product experimentation; * Experience with data warehouse/lake and orchestration (Airflow, Prefect, Dagster); **Languages** * Advanced English (reading documentation and research papers). **Academic background** * Bachelor’s degree in Computer Science, Engineering, Statistics, Mathematics, or related fields. **Soft skills** **Communication and collaboration** * Strong communication skills to clearly explain technical analyses and results to non-technical audiences. * Ability to collaborate effectively within multidisciplinary teams (data, product, engineering). **Learning and growth** * Intellectual curiosity and continuous learning—especially in generative AI and agents. * Openness to feedback and ongoing technical development. **Organization and accountability** * Attention to detail and rigor regarding data and documentation quality. * Ability to follow processes and best practices (versioning, code review, documentation). **Professional attitude** * Proactivity in identifying improvements to data, pipelines, and evaluations [whenever possible]. * Resilience and patience to handle experimentation, errors, and iteration. * Ethical stance in the use of data and AI models. * Ability to lead small initiatives and onboard new team members. **Additional information** Hybrid work opportunity in **Recife-PE**. **What we expect from all \#SiDiers:** Embrace the new Maintain open and direct dialogue Continuously develop Always do your best Build in partnership Think big, move fast **What we offer?** * **Work-life balance:** 40-hour weekly schedule under CLT regime, flexible hours with hybrid work (4 days in the office, 1 day remote); * **Well-being:** Gympass (WellHub, workplace gymnastics, quick massage, psychological support); * **Health:** Medical and dental insurance for you and your family; * **Children:** Daycare allowance for miniSiDiers, 120-day maternity leave, extended paternity leave; * **Future:** SiDi co-invests with you in private pension plans; * **Education:** Incentive program for continued studies and specialization; language fluency support; weekly talks on global *trend topics*; * **Meals:** Flexible meal and food vouchers; * **Transportation:** Transportation subsidy/cost assistance to reach SiDi, plus parking for those working onsite; * **Recognition:** Annual performance bonus and awards for SiDiers who’ve done something extraordinary; * **Diversity:** Committees focused on Well-being, Diversity, Mental Health, Social Impact, Sustainability, and Women in Tech; * **Modern offices:** Relaxed, collaborative environments with communal spaces, decompression rooms, pantries, and coffee machines; * **Still want more?** Dozens of partnerships offering benefits and discounts! For us, inclusion isn’t just about adding one more person to our team—it’s about embracing differences and unique perspectives. It is precisely this blend of techniques, skills, and stories that inspires us to align diverse viewpoints toward the same horizon—always moving forward. We operate in an environment of equal opportunity, regardless of gender, sexual orientation, religion, or disability. **All our positions are open to people with disabilities (PCDs).** JOB CODE: 5648 - N2 **We are one of Brazil’s largest science and technology institutes—and with a growing team of over 800 SiDiers, we’re already present in Campinas, Manaus, and Recife—the largest Brazilian technology and innovation parks.** Because those who aspire to build projects that will transform the world cannot stop transforming themselves. Over **20 years of history**, we have specialized in solving problems and carry in our portfolio **over 1,100 projects**, impacting millions of lives—driving innovation and making the future happen now.