Offer summary

Qualifications:

Strong experience with LLMs, NLP, and transformer-based architectures., Hands-on experience in fine-tuning and optimizing models using frameworks like Hugging Face, PyTorch, or TensorFlow., Proficiency in Python and familiarity with vector databases and retrieval-augmented generation., Strong background in statistical modeling and machine learning algorithms..

Key responsabilities:

Develop and fine-tune open-source LLMs for insurance-specific applications.

Optimize AI models to reduce hallucinations and increase accuracy.

Implement dynamic prompt engineering and integrate LLMs into a structured AI pipeline.

Research and deploy NLP-based solutions for text classification and entity recognition.

Job description

We are looking for a hands-on Generative AI Engineer who is deeply familiar with LLMs (Zephyr, Mistral, Llama, etc.), fine-tuning large models, and optimizing AI performance for industry-specific use cases. This person should have a strong engineering mindset and be capable of designing, training, and fine-tuning models while ensuring minimal hallucinations and high accuracy. The ideal candidate should also be well-versed in dynamic prompt engineering, retrieval-augmented generation (RAG), and knowledge base development.

Key Responsibilities:

Develop and fine-tune open-source LLMs (Mistral, Zephyr, Llama) for insurance-specific applications.
Optimize AI models to reduce hallucinations and increase domain-specific accuracy (insurance preferred).
Implement dynamic prompt engineering to control model responses and improve reliability.
Integrate LLMs into a structured AI pipeline (RAG, knowledge retrieval, vector databases).
Research, evaluate, and deploy NLP-based solutions for text classification, entity recognition, and semantic search.
Ensure models adhere to enterprise standards of accuracy, compliance, and security (GDPR, CCPA, PII handling).
Work with large datasets to train models efficiently and integrate them into real-world applications.
Optimize model inference efficiency for deployment in cloud-based and on-prem environments.
Collaborate with DevOps engineers to scale AI solutions in production.

Must-Have Skills & Experience:

Strong experience with LLMs, NLP, and transformer-based architectures.
Hands-on experience in fine-tuning and optimizing models (Hugging Face, PyTorch, TensorFlow, LangChain, etc.).
Proficiency in Python, PyTorch, TensorFlow, or JAX.
Familiarity with vector databases (Weaviate, Pinecone, FAISS) and retrieval-augmented generation (RAG).
Understanding of prompt engineering strategies to minimize hallucinations.
Strong background in statistical modeling and machine learning algorithms.
Knowledge of MLOps tools for deployment and monitoring (MLflow, Kubeflow, Ray).

We Offer

A dynamic and creative work environment with a team of passionate professionals.
Opportunities for professional growth and development.
Competitive salary and benefits package.
Flexible working hours and the possibility for remote work.

Why working at Diffco?

You will have the exciting opportunity to work on cutting-edge projects in the Silicon Valley for both US and European clients utilizing the latest technologies, methodologies, frameworks and approaches with the ability and time to learn and develop professionally.
Our team is a second family, so you would enjoy spending the day working on nice projects with kind people with broad interests.
Learn continuously, expand your skills and demonstrate your professional level and ability to take more responsibility to grow.
We care about you and are interested in your personal professional goals, motivation and we build a perfect work/life balance.

Required profile