This is us

At Avenga, we believe that human creativity empowers technology that matters. Operating globally, our 6000+ specialists provide a full spectrum of services, including business and tech advisory, enterprise solutions, CX, UX and Ul design, managed services, product development, and software development.

This is you

• Strong experience designing AI agents and orchestration flows, including tool-use, multi-step reasoning, and delegated logic.

• Hands‑on implementation of prototypes, orchestration logic, and interaction flows (Python / Node.js / notebooks / sandbox tools).

• Proficiency with agentic/LLM frameworks (e.g., Semantic Kernel, LangChain, AutoGen, or similar).

• Solid knowledge of prompt engineering and model behaviour tuning.

• Understanding of retrieval-augmented generation (RAG/CAG) and embedding/vector search principles.

• Ability to define small test sets, run structured checks, and interpret results for actionable improvements.

• Solid grasp of security, privacy, and risk considerations specific to AI (prompt/data handling, AI model security).

Nice to have:

• Familiarity with Azure AI components (Azure OpenAI, Cognitive Search, APIM, Service Bus/Event Grid).

Soft skills:

• Strong conceptual and analytical thinking; ability to translate ideas into practical prototypes and agent behaviours.

• Strong collaboration skills and clear technical communication.

• Curiosity, initiative, and comfort with exploratory work.

Preferred Qualifications:

• Prior experience with agent-based systems or multi-agent orchestration.

• Experience designing POCs or experimental architectures.

• Familiarity with cloud-native AI environments.

• Experience integrating AI into document, case management, or data‑centric platforms.

• Background in process automation (RPA, workflow engines, event‑driven architectures).

• Proficiency in English and Czech.

This is your role
1. Agent & Orchestration Design

• Design agent behaviours, reasoning flows, interaction patterns, delegation rules, and tool‑use strategies:

• Develop orchestration logic for multi-step, multi-agent, and tool‑augmented workflows.

• Recommend the optimal agentic patterns (planner–executor, cooperative agents, routing agents, etc.) for specific use cases.

• Collaborate with AI / IT architects to ensure agent and orchestration designs align with broader solution architecture.

1. Prompting & Behaviour Engineering

• Create and refine prompts, system instructions, behavioural frameworks, validation rules, and safety guidelines.

• Experiment with prompting techniques to improve consistency, reasoning quality, and task effectiveness.

• Contribute to a structured library of prompts and reusable behavioural templates.

• Ensure secure prompting for AI model users

• Help to raise awareness about the safe use of AI tools.

3. Agent & Orchestration Prototyping and Evaluation

• Build lightweight prototypes and proof‑of‑concepts to validate agent logic, orchestration patterns, and LLM capabilities.

• Compare alternative agent configurations, prompting strategies, and orchestration frameworks; document key design decisions and constraints.

• Document findings and provide clear recommendations to architects and development teams.

4. Evaluation & Quality Assurance

• Run structured evaluations of agent and workflow behaviour (reasoning consistency, tool‑use correctness, stability), including light regression checks.

• Detect and document hallucinations, edge cases, and failure patterns; propose mitigations and improvements.

• Prepare small, focused test sets or scenarios and interpret results to inform design and implementation.

• Provide clear, actionable recommendations to improve prompts, agent logic, retrieval parameters, and interaction patterns.

1. RAG/CAG Prototype Development

• Build conceptual and early-stage RAG/CAG pipelines focused on:

  • retrieval relevance

  • chunking logic

  • lightweight embedding strategies

  • prototyping of context assembly

• Evaluate retrievel quality to inform production-level architecture decisions.

6. Engineering Partnership & Pre‑Implementation Prototyping

• Partner with delivery engineers to review designs/code and challenge choices for feasibility, maintainability, and architectural alignment.

• Write minimal, purpose‑built prototype code to validate agent logic, orchestration, prompting strategies, and retrieval settings before platform implementation; provide reference snippets and interface notes to guide production work.

• Focus is on validation and enablement, not a core application development role.

7. Technical Collaboration & Knowledge Sharing

• Partner with AI/IT Architects, platform engineers, and business teams to refine requirements and technical direction.

• Share best practices on agent design, prompting strategies, orchestration logic, and retrieval concepts.

• Support development teams in adopting best practices in agentic and LLM-based design.

8. Continuous Learning & Internal Enablement

• Monitor trends in agentic architectures, orchestration frameworks, retrieval systems, and evaluation methods.

• Help define internal standards, reusable components, and design patterns for AI-driven solutions.

• Contribute to capability building across teams by sharing insights and prototypes.

What awaits you at Avenga?

• Pleasant working environment and stable background in one of the largest software development companies

• Space to work independently and creatively with the possibility to implement your own ideas and opportunities to learn new things and approaches

• You have the opportunity to influence your direction in the company

• High quality hardware and software, access to the latest technologies and specific projects

• Opportunities to improve your language skills, we are a major global player

• At Avenga, you'll get quality support in your development thanks to a wide range of internal and external training courses, so you can push the boundaries of your competences

• We offer you benefits such as life insurance and pension contribution, EN/DE/ES lessons, MultiSport card, sick days, meal allowance, lots of company events, offices in Prague Palmovka with quality amenities for lovers of sport, food and coffee, hybrid remote work after mutual agreement, and more
  
  At Avenga, everyone matters. We provide equal opportunities in recruitment, career development, and leadership, regardless of race, ethnicity, gender identity, sexual orientation, disability, age, religion, or any other characteristic. We are committed to fostering a work environment where our diverse community of employees, candidates, and business partners actively shapes our growth. By bringing together people from different backgrounds and experiences, we build a workplace where everyone feels free to be themselves while honoring the boundaries of others.