Flexible hybrid work (3 days remote, 2 days office)
Flexible working hours (start between 7:30-10:00)
We are developing next-generation digital products that bring AI capabilities into production, and we deliver applied AI solutions directly to enterprise clients. There are two directions: building new intelligent features from scratch, improving and automating existing processes, and implementing AI-driven solutions tailored to specific business domains and client needs. In this role, you will be the person who makes AI work in practice. Not every problem needs a large language model, sometimes the right answer is a well-tuned Random Forest, a statistical model, or a lightweight neural network. You will evaluate each problem on its own terms and choose the right approach: from classical ML and statistical methods through fine-tuned LLMs to agent-based systems and RAG pipelines. Your focus is on delivery: taking a business need and turning it into a working, deployed AI capability. You will split your time between two tracks: contributing to the AI layer of our digital products (data intelligence, intelligent document processing, data anonymization, business intelligence), and working on consulting engagements where you design and implement AI solutions for enterprise clients in e-commerce, manufacturing, healthcare, finance, and other domains. You will work within a cross-functional agile team alongside software engineers and product developers.
Key Responsibilities
Applied ML & Model Adaptation
Evaluate client and product problems to select the right ML approach: classical methods, lightweight deep learning, or LLM-based solutions
Build and deploy classical ML pipelines for tabular data, time series, classification, and regression problems
Design and implement end-to-end machine learning workflows, including data preprocessing, feature engineering, model development, and evaluation
Fine-tune and adapt pre-trained models to specific business domains and client requirements using LoRA, QLoRA, PEFT, and other parameter-efficient methods
Apply reinforcement learning and alignment techniques (RLHF, DPO) to optimize models for specific use cases
Design and implement RAG pipelines: embedding strategies, chunking, retrieval optimization, and vector database integration
Build and deploy LLM-based agents and autonomous workflows for both products and client solutions
Develop systematic prompt engineering strategies and evaluation frameworks for production AI systems
Deployment & Optimization
Deploy and serve ML models in production environments with focus on reliability, latency, and observability
Optimize inference costs through model selection, quantization, caching strategies, and right-sizing models to problem complexity
Implement model monitoring and evaluation pipelines for production systems: performance tracking, drift detection, hallucination assessment
Consulting & Product Delivery
Work directly with enterprise clients to understand their business problems and design practical AI solutions
Deliver rapid prototypes and PoCs to validate AI approaches before full implementation
Contribute AI features to our digital products: intelligent document processing, data anonymization, business intelligence, and music AI
Take ownership of ML features end-to-end: from business requirement analysis through implementation, deployment, and iteration
Required Qualifications
3+ years of professional experience in applied Machine Learning or ML engineering
Degree in Computer Science, Mathematics, Physics, or related fields (PMF, FER, FOI, or similar)
Proficient in Python and experience with ML/AI frameworks (PyTorch, TensorFlow, Hugging Face Transformers)
Experience with design and implement end-to-end machine learning workflows, including data preprocessing, feature engineering, model development, and evaluation
Understanding of classical ML methods: Random Forests, gradient boosting (XGBoost, LightGBM), SVMs, and statistical modeling, and knowing when to use them over deep learning
Familiarity with fine-tuning methods: LoRA, PEFT, adapter methods, transfer learning
Ability to understand business problems and translate them into practical ML solutions, choosing the right approach, from simple statistics to LLMs
Experience with deployment workflows
Proficiency with Git and collaborative development workflows
Excellent communication skills, including the ability to explain technical concepts to non-technical stakeholders
Preferred Qualifications
Solid prompt engineering skills: few-shot, chain-of-thought, systematic evaluation of prompts
Experience with reinforcement learning from human feedback (RLHF) and direct preference optimization (DPO) for model alignment
Experience with LLM integration, agent frameworks (LangChain or similar) or building and deploying RAG pipelines
Experience with model quantization and optimization: GPTQ, AWQ, GGUF, pruning, knowledge distillation
Experience with model evaluation for business use cases: domain-specific metrics, cost-benefit analysis, hallucination rate assessment
Experience with MLOps and model serving in production: vLLM, Triton, model monitoring, A/B testing of models
Experience with inference cost optimization: model selection for problem complexity, caching strategies, cost management
Experience with domain adaptation, tailoring models to specific industries (e-commerce, healthcare, finance, legal)
Experience in consulting or client-facing environments with rapid prototyping and PoC delivery
Familiarity with vector databases (Pinecone, Weaviate, Milvus, Chroma, or pgvector)
Understanding of CI/CD pipelines
