AI/ML Developer
We’re looking for a senior-level engineer who can design, build, and integrate advanced AI-driven solutions into real business workflows. This is a hands-on role that combines applied machine learning, systems integration, and high-performance software development.
What you’ll need to bring:
- At least 5 years of proven track record in building and deploying ML-based systems in production environments.
- English at Upper-Intermediate or Advanced level. You’ll be communicating with both technical and non-technical stakeholders.
- In-depth experience with LangChain or LlamaIndex, especially when it comes to tool integration and long-term agent memory.
- High-level Python skills, including the ability to write asynchronous code and design clean, maintainable APIs.
- Strong background working with vector databases (Pinecone, Weaviate, Milvus, Chroma or similar) and optimizing them for performance.
- Hands-on practice with embedding models
- A knack for designing and scaling knowledge graphs that handle large and complex datasets.
- Understanding of Model Control Plane (MCP) and its use for orchestrating LLMs while integrating with enterprise systems.
- Experience fine-tuning semantic search and architecting efficient RAG based solutions.
What you’ll be doing:
- Designing indexing pipelines that allow enterprise documentation (BRDs, BSDs) to be semantically searchable.
- Connecting to both private and public LLM endpoints, making sure performance, cost, and security stay in balance.
- Building systems to track agent context over time so they can remember past interactions.
- Developing multi-modal agent capabilities for tasks like requirement extraction, automated doc generation, meeting summaries, and Jira ticket creation.
- Creating a routing mechanism that intelligently picks the right LLM (GPT, Claude, Mistral, etc.) for each request.
- Setting up vector search solutions such as Weaviate or pgvector for fast contextual retrieval.
- Implementing privacy controls that ensure sensitive internal data never leaks to external systems.
- Designing prompt templates and evaluation flows that compare responses from different models.
- Embedding LLM powered capabilities into like Confluence, Jira.
- Introducing feedback loops through reinforcement learning or supervised fine-tuning so the agents keep improving.
- Building metrics dashboards and monitoring systems to track agent quality and model accuracy in real time.
- Working closely with backend/frontend teams to expose these AI features through APIs.
- Creating audit logs, usage records, and access controls to keep everything compliant.
- Participating in the broader system architecture design so AI components fit smoothly into the larger platform.
Frequently Asked Questions:
Does the job come with a probation period?
Yes, there is a three-month probation period.
What is the expected work schedule?
Full-time, flexible, but please start no later than 11:00 AM (average 8 hours per day). The Kyiv office is open from 9:00 AM to 8:00 PM. You can work remotely or choose a hybrid mode.
How many vacation and sick days are provided?
- Annual paid vacation – 21 working days;
- Paid confirmed sick leave;
- 6 unconfirmed sick days/days off per year.
What benefits are offered?
- English courses (up to $80/month after the probation period);
- Compensation for a sports center subscription (up to $80/month after the probation period);
- Medical insurance (50% after the probation period) or $200 per year for healthcare;
- Corporate events.
Hiring Process:
- HR call with a recruiter;
- Interview with the hiring manager (1.5 hours);
- Reference check;
- Offer for cooperation with Develux.
We believe that people with different perspectives and experiences drive innovation, helping us create unique solutions.
Note: Feedback regarding CV review results will be provided only if we decide to move forward with your application. Otherwise, your data will be retained in our database, and we will contact you if a suitable opportunity arises. The review process takes up to 10 working days.