Тип вакансії: Повна зайнятість

Розташування:

Automat-it is an all-in AWS Premier Partner and Managed Services Provider specializing in the startup ecosystem. With over 800 customers and 500+ AWS certifications, Automat-it brings hands-on expertise in AI, DevOps, and FinOps to empower fast-paced startups to grow, deliver & win. Our customers save significant time-to-market and optimize their cloud performance and costs.

We work across EMEA and the US, fueling innovation and solving complex challenges daily. Join us to grow your skills, shape bold ideas, and help build the future of tech.

We're looking for a Senior Data Scientist to join our Data Science practice and work directly with customers on complex AI and Machine Learning projects on AWS.

This is not a traditional analytics or research-focused Data Science role. You'll work across classical ML, fine-tuning, GenAI, multimodal systems, and production ML on AWS, helping customers decide which technical approach actually fits their problem.

You'll own projects end-to-end: understand the customer's domain and data, evaluate different solution paths, define the architecture, implement the critical parts, and guide delivery through production.

Work location: remote from Ukraine

If you are interested in this opportunity, please submit your CV in English.

Key Responsibilities
Own Data Science projects end-to-end: Take customer problems from technical discovery through data analysis, solution design, experimentation, implementation, deployment, and production validation.
Choose the right technical approach: Evaluate whether a problem should be solved with prompt engineering, RAG, GenAI, agentic workflows, fine-tuning, smaller language models, classical ML, Computer Vision, recommendation systems, or custom model training.
Work deeply with data: Analyze and prepare customer datasets, identify data quality issues, create representative validation and golden datasets, and ensure the available data can support the chosen modelling approach.
Build and fine-tune ML models: Train, fine-tune, optimize, evaluate, and deploy models using Python, PyTorch/TensorFlow, SageMaker, and modern ML tooling.
Work with Generative AI: Build and evaluate GenAI solutions using Amazon Bedrock, RAG, prompt engineering, model selection, agentic workflows, and related AWS-native AI services.
Use SageMaker in production: Work with SageMaker Studio, training jobs, endpoints, pipelines, model registry, batch inference, monitoring, and other production ML capabilities.
Design production-ready solutions: Make architecture decisions across quality, latency, cost, scalability, maintainability, observability, and operational complexity.
Work directly with customers: Participate in technical discovery, workshops, architecture discussions, and delivery conversations with founders, CTOs, engineering teams, and technical stakeholders.
Challenge technical assumptions: Help customers avoid unnecessary complexity, explain trade-offs, costs and recommend simpler or more effective approaches when appropriate.
Lead through technical ownership: Independently own projects with minimal supervision and mentor less experienced Data Scientists and engineers when needed.
Collaborate across teams: Work closely with AI Engineers, MLOps, Data Engineering, DevOps, and Solution Architecture teams on customer solutions that cross multiple technical domains.

Requirements
5+ years of experience in Data Science, Machine Learning, Applied Science, or a closely related role.
Strong classical Machine Learning fundamentals and hands-on experience building production ML solutions.
Deep practical experience working with data, including data preparation, validation, feature engineering, dataset construction, and model evaluation.
Strong Python skills and hands-on experience with PyTorch and/or TensorFlow.
Strong practical experience with AWS SageMaker beyond notebook-level usage, including model training, deployment, inference, pipelines, or production operations.
Hands-on experience with Amazon Bedrock and modern Generative AI approaches.
Practical experience with fine-tuning models and understanding when fine-tuning is preferable to prompting, RAG, or other approaches.
Experience in at least one strong ML domain such as NLP, Computer Vision, recommendation systems, forecasting, structured ML, multimodal ML, or similar.
Understanding of RAG, embeddings, prompt engineering, foundation models, and agentic workflows.
Strong understanding of MLOps and production ML practices, including model deployment, monitoring, reproducibility, lifecycle management, and CI/CD.
Experience designing and owning solutions independently rather than working only from predefined technical specifications.
Strong customer-facing communication skills and ability to explain technical trade-offs clearly.
Ability to work with ambiguity, messy real-world data, and changing customer requirements.
Strong technical judgment and a pragmatic approach to balancing model quality with delivery speed, cost, and business value.

Nice to have
Experience with AgentCore, Bedrock Agents, LangGraph, Strands Agents, or other agentic frameworks.
Experience with Small Language Models or domain-specific model adaptation.
Experience with recommendation systems, audio ML, signal processing, or multimodal systems.
Experience in technical consulting, pre-sales, or customer discovery.
AWS Machine Learning certifications.
Master's degree or PhD in Computer Science, Machine Learning, Data Science, Mathematics, Statistics, or a related field.

What Success Looks Like in Year One
Independently owns assigned data projects end-to-end, from understanding the problem and making technical decisions to implementation, troubleshooting, testing and successful delivery, with minimal supervision.
Has built strong practical expertise across the team's core data stack and can confidently work across different customer environments, including Spark, Python/SQL, data lakes/lakehouses, analytical and NoSQL databases, streaming and other technologies used by the team.
Can investigate unfamiliar data problems independently, identify performance or data-modeling issues, understand upstream/downstream dependencies, and recommend pragmatic solutions to customers.
Continuously expands technical depth, keeps up with modern Data technologies and patterns, and can quickly learn new technologies as customer projects and the Data practice evolve.

Automat-it is committed to fostering a workplace that promotes equal opportunities for all. We firmly believe that cultivating a diverse workforce is crucial to our success. Our recruitment decisions are grounded in your experience and skills, recognizing the value you bring to our team.

The post Senior Data Scientist (SageMaker, Bedrock) appeared first on Remote Work Ukraine [Українське].