Job Summary: We are seeking a professional to develop and optimize Machine Learning models—focused on Transformers and LLMs—and build robust textual processing pipelines. Key Highlights: 1. Develop and optimize Machine Learning and LLM models. 2. Collaborate on technical decisions and scale solutions into production. 3. Lead by example, with a focus on results and innovation. ### **JOIN THE DOCKE TEAM** Transformation is part of our history and motivates us to keep moving forward—we are the protagonists of change. Being a Docke team member means having our values embedded in our DNA. At Docke, we act responsibly and urgently to deliver high-quality work. We empathize and move together toward the same goal. We respect differences and value diversity. We are fearless and make things happen. Each of us is prepared to overcome challenges and turn the page in constant pursuit of our best selves. That’s why we lead by example, act like owners, and remain results-oriented. If you, like us, believe in these same principles, this is your opportunity to be part of something bigger. Join this incredible journey of evolution. **Responsibilities and Duties** * Develop, train, and optimize Transformer- and LLM-based models for text processing tasks; * Build robust textual preprocessing and postprocessing pipelines; * Apply software engineering best practices in Machine Learning projects; * Work with structured and unstructured data, integrating models with databases and APIs; * Participate in technical decisions regarding architecture, model versioning, testing, and deployment; * Collaborate with data engineers and other teams to scale solutions into production. **Requirements and Qualifications** * **Python**: Advanced proficiency in the language, with emphasis on object-oriented programming, code optimization, and project organization; * **Machine Learning**: Experience with models such as BERT, RoBERTa, GPT, Mistral—including fine-tuning and application in NLP tasks. Strong experience in prompt engineering; * **Text Processing**: Classification and named entity extraction; * **OCR**: Basic experience with tools such as Tesseract and Google Vision OCR; * **Databases (PostgreSQL)**: Data manipulation via SQL, integration with ML pipelines; * **Git/GitHub**: Code versioning, PR reviews, usage in collaborative teams; * **Debugging and Testing**: Ability to identify and resolve bugs; create automated tests using Pytest or Unittest; * **Design Patterns**: Application of architectural best practices and design patterns. * **LLMs** — (GPT, Mistral, Claude - Sonnet, etc); **Preferred Qualifications** * Experience with RESTful APIs and FastAPI; * Knowledge of Docker and cloud-based model deployment; * Familiarity with prototyping tools such as Streamlit; * Knowledge of cloud platforms (AWS, GCP); * Experience with LLM-based Agents; * Use of MLflow, Weights & Biases, DVC, or similar tools for experiment monitoring, versioning, and automation; * **MLOps**, or Machine Learning Operations, refers to the set of practices and tools that automate and streamline the machine learning lifecycle—from development through deployment and model monitoring. **Additional Information** **Benefits:** At Docke, we know details make all the difference! Here, we offer an amazing benefits package—take a look: Meal and Food Allowance via Flash, for when hunger strikes. Health and dental insurance to take care of you. Life insurance for peace of mind. Pharmacy Benefit, to save and support your health. Petlove—because at Docke, we understand your furry family members matter too. Psicologia Viva, psychological support at your fingertips. Wellhub and TotalPass to keep your body moving. Edupass—for continuous learning and development! Sesc partnership, offering leisure and culture. Childcare allowance for mothers and fathers with children up to 5 years old. Baby Cash when your family grows. Birthday Day Off to celebrate as you deserve. Working hours: 44 hours per week At Docke, we value your well-being and growth. Join our team and enjoy these amazing benefits! Work model: Location: Pinheiros (hybrid model with 4 days onsite and 1 day remote) #### **JOIN THE NEW STANDARD FOR DOCUMENT MANAGEMENT** **We are Brazil’s largest legaltech, with technology connecting companies to +35,000 public agencies—making document retrieval and analysis faster, more secure, and more strategic.** More than a platform, we are the infrastructure behind operations—powering businesses with the market’s best user experience and fastest response time.