Enter projects early, before the exact solution is known, run technical workshops, talk with management about real problems, validate feasibility, and translate them into architecture
Define success metrics and the value to be delivered at each stage of an implementation
Build the solution yourself: agents, RAG systems, orchestration, and production code in Python and SQL, from rapid prototype to production
Integrate our product with customer APIs, data silos, and legacy systems inside their security constraints
Work with imperfect environments and organize fragmented data when there is no dedicated data team on the customer side
Build evals, baselines, quality gates, and monitoring that remain operational after the implementation is handed over
Design systems customers can rely on, including retries and model fallbacks, cost limits, human-in-the-loop for irreversible actions, prompt-injection resistance, and strict data boundaries
Drive adoption and measure it, making sure people actually use the solution and that the implementation produces visible results
Turn recurring implementation patterns into reusable modules, accelerators, and ready-made solutions that shorten future deployments
Bring product gaps and customer needs back to Product and Ops so implementation insights directly influence the roadmap
Support sales in meetings with new customers by demonstrating the product, designing the first use case, and defining how success will be measured
Work with tools such as Claude Code or Cursor while remaining accountable for the quality of the code that reaches the customer
Our requirements
4–6 years of experience in software engineering, including LLM-based solutions shipped to production rather than left at prototype stage
Very strong Python and SQL skills, production APIs, and integrations with systems customers already use
Strong knowledge of agent architectures: tool calling, context engineering, sub-agents, state management, and orchestration, including MCP servers or custom tool-calling interfaces
Ability to choose an agent architecture based on the task and its expected failure modes
Production experience with retrieval: vector databases, embeddings, hybrid search, re-ranking, chunking, and context management
Experience with data modeling, ETL/ELT pipelines, performance tuning, and unstructured data
Experience with evaluation systems for generative solutions, from baselines and regression tests to LLM-as-a-judge, plus tracing and observability in Langfuse or LangSmith
Experience deploying and maintaining cloud systems, especially Azure, with containerization and CI/CD
Understanding of application security: OWASP, secrets management, input validation, permissions, personal data, and attacks specific to LLM applications
Experience leading customer implementations from discovery to production and communicating technical trade-offs to non-technical decision-makers
Comfort working in immature environments with spreadsheet-based data, limited documentation, and no dedicated data team
Strong judgement when narrowing scope, setting priorities, and balancing speed, quality, and scope
Ability to define implementation success metrics and demonstrate business value in the customer’s own numbers
Daily use of agentic engineering tools and the ability to review and productionize model-generated code
Very good communication skills in Polish and English
Optional
Experience building this type of function in a company from scratch
Experience in a similar role in a product company
Knowledge of the systems mid-sized companies actually run on, including ERP, CRM, accounting systems, and spreadsheet-heavy workflows
Ability to quickly build a simple interface to demonstrate an idea instead of only describing it
Experience handing a solution over to a customer team so they can continue developing it independently
Experience in presales, technical demos, and proof-of-value design
Experience expanding a customer relationship from the first implementation into additional areas
What we offer
End-to-end ownership of customer implementations, from discovery and architecture to production and adoption
Direct work with customers, including management, technical teams, and real production environments
A hands-on role combining engineering, product discovery, and business problem solving
Real influence on the product roadmap through insights gathered during implementations
The opportunity to turn customer-specific solutions into reusable modules and accelerators
An AI-native, fast-moving environment with a strong focus on ownership, measurable outcomes, and continuous exploration