Expert LLM Engineer (AI Security)

Location: Hybrid in Gdańsk, Poland, with regular in-studio collaboration in Gdańsk, Wrzeszcz.

Salary: 200–240 PLN/hour + VAT on a B2B contract, or 24,500–29,400 PLN gross per month under an employment contract.

Company

hirify.global partners with ambitious brands to create digital products across strategy, UX, engineering, and long-term growth.

What you will do

• Build and scale production AI-powered security solutions for a passwordless authentication platform using advanced biometrics and serving millions of users worldwide.
• Develop backend components in Python that integrate multiple LLMs with applications, APIs, and data sources.
• Design and operate RAG and multi-agent systems, including agent orchestration, tools, integrations, context, memory, and agent fleet management.
• Evaluate LLM response quality and agent behaviour through evaluation datasets, automated testing, validation, logging, tracing, and traffic analysis.
• Investigate vulnerabilities and attack paths through threat modelling and controlled adversarial testing, including prompt injection, tool misuse, safeguard bypasses, and abuse in Android and iOS environments.
• Optimise performance and cost, deploy and scale LLM-based solutions, and translate security findings into stronger safeguards and product improvements.

Requirements

• At least 7 years of experience in software engineering, machine learning engineering, or security engineering, including at least 2 years developing and maintaining LLM-based solutions in production.
• Good knowledge of Python and experience building maintainable backend components integrating multiple LLMs with applications, APIs, and data sources.
• Experience designing multi-agent architectures, orchestrating agent collaboration, building agent tools, and managing agent fleets, context, and memory.
• Hands-on experience building RAG systems and working with embeddings and vector databases.
• Experience evaluating LLM response quality and agent behaviour, building evaluation datasets and automated testing pipelines, and understanding data quality, validation, and data leakage.
• Experience monitoring LLM and agent activity through logging, tracing, and traffic analysis to identify suspicious sources, malicious behaviour, and unauthorised activity.
• Experience optimising LLM solution performance and costs through model selection, caching, and rate-limit management.
• Practical experience investigating vulnerabilities in LLM applications and agentic systems through threat modelling and controlled adversarial testing, including prompt injection, tool misuse, and abuse protection.
• Experience deploying, operating, and scaling LLM-based solutions; security research or red teaming experience beyond LLMs; and familiarity with Docker, CI/CD pipelines, and cloud services.
• Very good communication skills in English and Polish.

Nice to have

• Familiarity with garak, promptfoo, or Langfuse.

Culture & Benefits

• Real ownership and influence over products, with a focus on business value over technical perfection.
• Continuous feedback, open communication, regular 1:1s, mentoring, internal workshops, and knowledge sharing.
• Flexible working hours, with start times between 7:00 and 10:00, in a hybrid model.
• High-quality equipment, including a MacBook and testing devices, plus an individual training budget.
• Private medical care through Compensa, MultiSport, English classes in small groups, and team gatherings.