Middle ML/GenAI Engineer (IRC269413)

Job Description

Must-Have Skills
- 3+ years of experience in Machine Learning, Generative AI, and/or NLP.
- Strong expertise with LLMs (e.g., GPT-4, T5, Claude, LLaMA) and generative AI techniques.
- Proficient in Python; hands-on experience with ML frameworks such as PyTorch and LangChain, and LangGraph.
- Experience with using prompt engineering techniques to adapt LLMs effectively and/or fine-tuning LLMs using APIs like OpenAI and HuggingFace.
- Strong grasp of data structures and data transformation processes.
- Experience serving ML models as API services in production environments.
- Strong knowledge of different types of databases (SQL, noSQL, Vector, etc).
- Upper-Intermediate English (comprehensive written and verbal communication).
- Fluent Ukrainian (comprehensive written and verbal communication).
- Be familiar with Agile methodologies and Scrum.

Nice-to-Have Skills
- Experience developing AI agents with external tool integration.
- Experience with natural language-to-SQL or natural language-to-code generation systems.
- Familiarity with AWS and its computing services.
- Background in HCI, data storytelling, or user-centric AI systems.

Job Responsibilities

- Use and fine-tune large language models (e.g., GPT, LLaMA, T5).
- Engineer and optimize prompts for LLM-based workflows.
- Implement and integrate retrieval-augmented generation (RAG) and tool-augmented agents.
- Build components of an agentic system capable of multi-step reasoning.
- Stay current with advancements in LLMs, agent frameworks, and RAG techniques.
- Research and propose improvements to model performance, usability, and tuning.
- Prototype ideas for intent recognition, query synthesis, analysis, and reasoning.
- Collaborate closely with developers and other stakeholders to ensure system alignment with product and business goals.

Department/Project Description

Working with a data platform that gathers information about the appropriate company from various sources, normalizes it
according to the predefined flow, and shows it to the user as analytical dashboards. The gathered and processed information
lets the end-user decide regarding the investigated company's business state. Analytical information helps to predict the
future state of the appropriate company.

As an investor, the client provides high-value analysis and support to partner companies to identify and mitigate emerging
business challenges. Today, the process is highly manual and fragmented.

The project aims to define and bring to life a digital platform that connects client employees with relevant and meaningful
information about their portfolio holdings, enabling insights and action.