Lead Developer — Agentic AI Platform

Описание на позицията

Purpose Own the technical implementation and framework development of the Agentic AI Platform, with a strong focus on Python-based platform components, reusable agentic AI patterns, and developer enablement. Ensure that the platform codebase is modular, maintainable, secure, extensible, and supports efficient delivery of agentic AI use cases across the organization.

Responsibilities Develop and maintain Python-based platform capabilities, reusable components, SDKs, templates, and developer tools.

Design and implement reusable framework capabilities for agents, tools, MCP servers, orchestration, memory, RAG, model access, and evaluation.

Implement MCP servers end-to-end, including schema design, validation logic, action execution, error handling, and interface contracts.

Extend agent runtime capabilities, including planning, tool-use, routing, reasoning support, context handling, and orchestration patterns.

Design and evolve RAG and knowledge access logic, including chunking, retrieval, ranking, generation, and evaluation patterns.

Implement agent memory and context management logic, including semantic, vector-based, and session-related memory concepts.

Build test frameworks, evaluation harnesses, and enablement tooling that allow developers to test, validate, and improve agentic AI solutions.

Ensure clean separation between shared platform/framework capabilities and use case-specific implementation logic.

Evaluate, integrate, and extend open-source AI and agentic AI frameworks in line with the platform architecture.

Drive code quality, modularity, refactoring, secure coding, documentation, and maintainability of the platform codebase.

Requirements / Skills Strong hands-on Python development experience.

Experience developing reusable frameworks, SDKs, libraries, platform components, or developer tools.

Deep understanding of open-source AI and agentic AI frameworks such as LangChain, LangGraph, Semantic Kernel,

AutoGen, LlamaIndex, CrewAI, or comparable frameworks.

Strong understanding of LLM-based agent architectures, including planning, tool use, routing, orchestration, reasoning patterns, and agent runtime design.

Hands-on experience with MCP-style servers, action or skill servers, schema design, validation logic, and execution interfaces.

Solid knowledge of RAG systems, including retrieval logic, chunking, ranking, generation, grounding, and evaluation.

Good understanding of agent memory concepts, including vector memory, semantic memory, session context, and long-term memory patterns.

Strong skills in API design, schema modeling, interface contracts, modular software design, and distributed systems.

Experience with test automation, evaluation harnesses, quality gates, CI/CD, secure coding, and code quality practices.

Familiarity with knowledge representation concepts such as ontologies, taxonomies, and knowledge graphs is beneficial.

Our Offer Attractive remuneration package with a full-time permanent contract

Yearly bonus

Incentives for high performance

Employee referral bonuses

Stable work environment part of large Banking and Insurance Multinational Group

Working on long-term in-house projects

Quiet office in Kamenitza Office Park

Flexible working hours

Remote work possible

Additional health insurance including dental care

Food vouchers

Multisport card

Exchange of experience and training with international professionals

Career development opportunities

Team building events held both in Bulgaria and Europe

A friendly team culture that promotes learning and team collaboration

German language courses

By submitting your application you acknowledge and consent to the data privacy policies of UNIQA Group. You grant us use of your personal data for the purpose of validating your professional qualifications for the purposes of potential employment by UNIQA Group Bulgaria branch.