Accountabilities:

• Independently own and execute AI research tasks or sub-projects, delivering high-quality experimental results and meaningful contributions to research initiatives.
• Design and implement complex experiments to validate hypotheses, define evaluation protocols, and compare model performance against relevant benchmarks.
• Analyze research outcomes, interpret results, and communicate clear conclusions and recommendations.
• Develop high-quality, reusable code contributions for machine learning research repositories and reference implementations.
• Improve internal benchmarks by identifying data gaps, proposing evaluation metrics, and introducing new models or comparison methods.
• Create data-driven research workflows to improve model training, evaluation, and performance measurement.
• Design and optimize pipelines for evaluating large language models, including quality, safety, faithfulness, coherence, and reasoning capabilities.
• Collaborate with technical and non-technical stakeholders by translating complex AI concepts into clear guidelines and actionable insights.
• Conduct controlled experiments, ablation studies, and adversarial testing to understand model behavior and improve reliability.
• Contribute to the development of scalable AI research practices through automation, documentation, and engineering excellence.

Requirements:

• PhD or Master’s degree in Computer Science, Computational Linguistics, Machine Learning, or a related quantitative field.
• 3+ years of hands-on experience in applied NLP research, machine learning engineering, AI research labs, or data-centric AI environments.
• Strong theoretical and practical knowledge of Transformer architectures, including GPT-style decoder models, encoder-decoder architectures, attention mechanisms, positional embeddings, and tokenization techniques such as BPE or SentencePiece.
• Extensive experience with post-training methods, including Supervised Fine-Tuning (SFT), RLHF, PPO, and Direct Preference Optimization (DPO).
• Experience working with large-scale datasets, noisy label handling, data annotation strategies, active learning, and semantic bias analysis.
• Knowledge of LLM evaluation approaches, including LLM-as-a-Judge pipelines, pairwise comparison systems, and reference-free metrics.
• Understanding of AI agent evaluation, including function/tool calling, ReAct frameworks, and reasoning trajectory assessment.
• Experience designing data evolution pipelines, knowledge distillation techniques, and approaches to reduce model collapse from synthetic data.
• Familiarity with adversarial testing, safety evaluation, prompt stress testing, and balancing helpfulness with model safety.
• Expert-level Python skills and strong experience with deep learning frameworks such as PyTorch, TensorFlow, or JAX.
• Ability to process and manage large-scale text datasets using scripting, JSONL manipulation, regular expressions, and data optimization techniques.
• Experience with data versioning tools such as DVC or LakeFS is a plus.
• Strong analytical thinking, research communication skills, and ability to work effectively in ambiguous environments.

Benefits:

• Fully remote work opportunity for candidates based in Brazil.
• Flexible work arrangements with opportunities for eligible candidates to access office locations.
• Opportunity to contribute to cutting-edge AI research and large-scale machine learning projects.
• Exposure to advanced technologies in NLP, generative AI, model evaluation, and data engineering.
• Collaboration with global teams of AI researchers, engineers, and technical specialists.
• Opportunity to influence the development of innovative AI systems and research methodologies.
• Professional growth through complex research challenges and continuous learning opportunities.
• Inclusive and collaborative environment focused on innovation, diversity, and technical excellence.