AI Engineer

Business At Work; Paris; Sur place; EUR 70 000 - 110 000; Plein temps

Résumé du poste

Business At Work is seeking an AI Engineer as a full-time subcontractor/freelancer to design and scale intelligent workflows for a cybersecurity SaaS platform. You will build AI-powered use cases that augment compliance evaluation workflows and design agents within a custom internal framework to process structured data end-to-end.

You will collaborate with a Cross-functional team including a Full Stack Engineer, Cybersecurity Standards Engineer, UI/UX Designer, and Product Manager to define data

Qualifications

• 3-4 years of professional software development experience
• Strong JSON data structures understanding
• Prompt engineering experience is a plus
• Familiarity with LLM-based agents and evaluation/benchmarking is a plus
• English proficiency required; French/German a plus

Responsabilités

• Design and implement AI agents using our internal framework for compliance workflows on the platform.
• Build agents that process structured JSON data at all evaluation stages: pre-, in-, post-evaluation.
• Develop, iterate, and maintain prompts for reliable, well-structured outputs from AI models.
• Collaborate with the Cybersecurity Standards Engineer to use correct data structure and semantics.
• Integrate agents into backend and client library with the Full Stack Engineer.
• Work with the UI/UX Designer to present AI outputs clearly and credibly in the frontend.
• Contribute to the internal agent framework by identifying reusable blocks across use cases.
• Document agent designs, prompt strategies, and data flow assumptions for the team.
• Identify new AI use cases in collaboration with the Product Manager.

Description du poste

About the Role

We are looking for an AI Engineer acting as a full-time subcontractor/freelancer to design and scale intelligent workflows in close collaboration with our internal team. Our product is a cybersecurity SaaS platform that helps manufacturers and security teams certify IoT and connected devices against international regulatory standards. It is built around three components: a React/TypeScript frontend, a TypeScript client library, and a Node.js/PostgreSQL backend, with AI-assisted features that guide users through compliance workflows. In this role, you will design and implement AI-powered use cases that augment the platform's compliance evaluation workflows. You will work within a custom internal agent framework to build agents that process structured platform data before, during, and after compliance evaluation flows, helping users navigate complex certification requirements more efficiently. You will work closely with the Full Stack Engineer to integrate agents into the platform, with the Cybersecurity Standards Engineer to understand the compliance data those agents operate on, with the UI/UX Designer to surface AI outputs clearly and trustworthily in the frontend, and with the Product Manager to identify and scope new AI use cases.

Key Responsibilities

• Design and implement AI agents using our custom internal agent framework, tailored to compliance evaluation workflows on the platform.
• Build agents that process structured platform data (JSON) at each stage of an evaluation flow: pre-evaluation preparation, in-flow guidance, and post-evaluation analysis and reporting.
• Develop, iterate, and maintain prompts that reliably produce accurate, structured, and contextually appropriate outputs from AI models.
• Work closely with the Cybersecurity Standards Engineer to understand the structure and semantics of compliance rule sets and evaluation data, and ensure agents operate on that data correctly.
• Collaborate with the Full Stack Engineer to integrate agents cleanly into the backend and client library.
• Partner with the UI/UX Designer to define how AI outputs are presented, explained, and actioned in the frontend.
• Contribute to the internal agent framework by identifying gaps, patterns, and reusable building blocks across use cases.
• Document agent designs, prompt strategies, and data flow assumptions clearly for the broader team.
• Proactively identify new AI use cases within the platform in collaboration with the Product Manager.

Requirements

• 3-4 years of professional software development or technical experience.
• Strong, hands-on understanding of JSON data structures: comfortable reading, designing, and reasoning about complex nested data.
• Experience working with structured output techniques (JSON mode, function calling, schema validation) is a plus.
• Demonstrated experience with prompt engineering: crafting, testing, and iterating prompts to produce reliable and well-structured outputs from large language models.
• Exposure to evaluation and benchmarking strategies for AI-generated outputs is a plus.
• Familiarity with how LLM-based agents work in practice: context management, tool use, chaining, and output validation.
• Experience working with agent frameworks (LangChain, LlamaIndex, custom frameworks, or similar) is a plus.
• Methodical and systematic approach to problem-solving: able to break down a complex workflow into discrete, testable steps.
• Ability to read and understand TypeScript code well enough to collaborate effectively with the engineering team, familiarity with TypeScript at working level is a plus.
• Clear written communication skills for documenting prompt logic, agent designs, and data contracts.
• English is a must; French or German is a plus.

Soft skills

• Detail-oriented and rigorous.
• Curious and self-directed.
• Collaborative and communicative.
• Comfortable with iteration.