• Develop, fine-tune, and evaluate LLMs, diffusion models, VLMs, and transformer-based architectures.

• Train models using custom datasets, including text, image, audio, and multimodal data.

• Implement prompt engineering, instruction tuning, and retrieval-augmented generation (RAG) pipelines.

AI System Engineering

• Build scalable inference pipelines using Python and modern ML frameworks (PyTorch, HuggingFace, TensorRT, ONNX).

• Implement model optimization techniques (quantization, distillation, pruning) for deployment on cloud or edge devices.

• Integrate generative AI models into backend microservices and APIs.

Data Engineering & Preparation

• Create robust data pipelines for dataset collection, preprocessing, augmentation, and evaluation.

• Maintain dataset quality with labeling, cleaning, filtering, and benchmarking.

Research & Rapid Experimentation

• Explore new architectures, training techniques, and SOTA research papers.

• Run controlled experiments, A/B tests, and performance validations.

• Prototype and iterate quickly to convert ideas into functional demos.

Production Deployment

• Work with DevOps/MLOps teams to package and deploy models in production (Docker, Kubernetes preferred).

• Monitor model performance, drift, quality, latency, and reliability once deployed.

Required Skills & Experience

• 1–3 years experience in applied ML or generative AI is a MUST

• Strong proficiency in Python , PyTorch/TensorFlow, and HuggingFace ecosystem.

• Practical experience with LLMs, diffusion models, or vision transformers .

• Hands-on experience fine-tuning models on custom datasets.

• Familiarity with cloud services (AWS/GCP/Azure) or on-prem GPU stacks.

• Understanding of transformer architecture, embeddings, tokenization, attention mechanisms.

• Experience with APIs, microservices, or backend engineering using Python (FastAPI/Flask).

• Strong debugging, experimentation, and documentation skills.

Nice-to-Have

• Experience with OpenAI, Anthropic, Mistral, Llama, Qwen , or similar model families.

• Familiarity with vector databases (Pinecone, Weaviate, FAISS).

• Knowledge of RAG , agentic workflows , and LLM tool-use frameworks (LangChain, LlamaIndex).

• Exposure to multimodal models (e.g., CLIP, VLMs, SAM, diffusion models).

• Experience in medical imaging, radiology AI, or healthcare AI (bonus for Biocliq/Eliqsar context).

• Contributions to open-source ML projects or published research.

Education

• Bachelor’s or Master’s degree in Computer Science, AI/ML, Data Science, Applied Math, or a related field is a MUST.

• Equivalent practical experience is also considered.