PROMPT ENGINEER

Ayesa Digital; Brussel Hoofdstad; Sur place; EUR 70 000 - 110 000; Plein temps

Résumé du poste

Ayesa Digital is seeking a Prompt Engineer to design, develop and iterate sophisticated prompt strategies for Large Language Models to solve complex business problems. You will architect prompts, implement RAG systems, curate knowledge bases, and fine-tune models using PEFT techniques.

Join an international team delivering high-impact EU projects with a focus on ethics, safety, and responsible AI practices, in a growing global company.

Qualifications

• Deep understanding of NLP fundamentals and LLM operation.
• Mastery of prompting techniques including Zero-shot, Few-shot, CoT, Self-Consistency, Tree-of-Thoughts.
• End-to-end development of RAG systems with vector DBs and embedding models.
• Hands-on LLM fine-tuning (LoRA/QLoRA) and hyperparameter tuning.
• Proficiency in Python for scripting and AI API integration.
• Experience with LLM evaluation frameworks and metrics.
• Familiarity with LangChain/LlamaIndex for building apps.
• Knowledge of cloud AI platforms for deployment and management.
• Strong software engineering fundamentals for scalable AI solutions.
• Understanding of AI ethics and cybersecurity concerns for LLMs.

Responsabilités

• Architect, test, and refine prompts through experiments for accuracy and cost-efficiency.
• Develop and implement RAG systems grounded in a proprietary knowledge base.
• Build and curate knowledge bases and datasets for RAG and fine-tuning.
• Fine-tune LLMs using PEFT techniques for domain-specific use cases.
• Create evaluation frameworks to assess AI model and prompt performance.
• Design and implement complex prompt chains and automation workflows.
• Analyze model failures and mitigate issues like hallucination and prompt injection.
• Create guardrails and content validation to ensure safety and compliance.
• Stay updated with generative AI advances and best practices.

Connaissances

NLP fundamentals

LLMs operation

Prompt engineering

RAG systems

Vector databases

LoRA/QLoRA

Python scripting

LangChain

LlamaIndex

Cloud platforms

Ethics & security

Outils

Pinecone

Weaviate

pgvector

OpenAI API

Azure OpenAI

Hugging Face

LangSmith

MLflow

Description du poste

At Ayesa Digital, we grow with you.

Every professional in our company is essential. Thanks to their talent, we continue to expand: today, we are a global team of more than 11,000 people working toward a shared mission.

Ayesa Digital is currently participating in high-impact European Union projects designed to address major European challenges and drive science and innovation. These strategic technological initiatives stand out for their international scope and strong commitment to socially oriented results.

We are looking for a Prompt Engineer to join our international team and contribute in the design, develop and iterate o sophisticated prompt strategies for Large Language Models (LLMs) to solve complex business problems.

Key Responsibilities

• Architect, test, and refine prompts through systematic experimentation to optimize for accuracy, relevance, creativity, and cost-efficiency.
• Develop and implement Retrieval-Augmented Generation (RAG) systems to ground LLM responses in a specific, proprietary knowledge base.
• Build, curate, and maintain high-quality knowledge bases and datasets from existing documentation for use in RAG and fine-tuning.
• Fine-tune and adapt pre-trained LLMs using techniques like Parameter-Efficient Fine Tuning (PEFT) for specific domains and use cases.
• Develop and implement evaluation frameworks and metrics to quantitatively and qualitatively assess AI model and prompt performance.
• Design and build complex prompt chains and workflows to automate multi-step tasks and business processes.
• Analyze model failures and implement mitigation strategies for issues like hallucination, bias, and prompt injection attacks.
• Create robust guardrails and output validation systems to ensure AI-generated content is safe, ethical, and compliant.
• Stay current with the latest advancements in generative AI and prompt engineering techniques to ensure best practices.

What We Are Looking For (Requirements):

• Deep understanding of Natural Language Processing (NLP) fundamentals and the architecture/operation of Large Language Models (LLMs).
• Mastery of advanced prompting techniques: Zero-shot, Few-shot, Chain-of-Thought (CoT), Self-Consistency, and Tree-of-Thoughts prompting.
• Proven experience with the end-to-end development of RAG systems, including vector databases (e.g., Pinecone, Weaviate, pgvector), embedding models, and retrieval strategies.
• Hands-on experience with LLM fine-tuning techniques (e.g., LoRA, QLoRA) and hyperparameter tuning (temperature, top-k, top-p).
• Proficiency in Python for scripting, data manipulation, and interacting with AI APIs (OpenAI, Anthropic, Azure AI, Hugging Face, etc.).
• Experience with LLM evaluation frameworks and metrics (e.g., using tools like LangSmith, MLflow, or custom pipelines).
• Familiarity with AI orchestration frameworks such as LangChain and LlamaIndex for building complex applications.
• Knowledge of cloud AI platforms (e.g., AWS Bedrock, Google Vertex AI, Azure OpenAI Service) for model deployment and management.
• A firm grasp of data structures, algorithms, and software engineering principles for building scalable and maintainable AI solutions.
• Understanding of AI ethics, responsible AI principles, and cybersecurity threats specific to LLMs (e.g., data poisoning, prompt injection).

What We Offer:

• Prestigious projects within European institutions.International, innovative, and multicultural environments.Continuous support from a team of experts in EU projects.

  If you are ambitious, enthusiastic, and seeking a new professional challenge in international projects with real-world impact, this is the place for you!

  In accordance with Organic Law 3/2007 of March 22, the company is committed to promoting the defense and effective application of the principle of equality between men and women, preventing any type of labor discrimination based on sex, and guaranteeing equal entry opportunities. Furthermore, we promote diversity and reject any discrimination based on race, gender, functional diversity, religion, sexual orientation, gender identity, or any other personal or social condition, striving to build an inclusive and enriching environment.