Key Responsibilities
- Develop reusable AI/ML models and end-to-end data science assets / products.
- Build and deploy LLM-based applications, including fine-tuning, prompt engineering, and RAGs.
- Translate domain challenges into data science problems with actionable solutions.
- Perform exploratory data analysis (EDA) to design optimal approaches.
- Rapidly prototype and iterate using ML/DL and LLMs.
- Deploy and manage models in Azure cloud environments.
- Mentor junior team members to foster growth and ensure quality deliverables.
LLM Tools & Frameworks
- Pretrained LLMs: OpenAI (GPT series), Gemini, LLaMMa, DALL-E, Hugging Face Transformers, T5, BERT, RoBERTa, etc.
- Fine-tuning Frameworks: Hugging Face Trainer, LoRA, PEFT etc.,
- Prompt Engineering Tools: LangChain, LangGraph, Semantic Kernel, Guidance etc.,
- Vector Databases: Pinecone, Weaviate, FAISS, Chroma etc.,
- LLM Evaluation Tools: OpenAI Evaluation Toolkit, Language Model Evaluation Harness etc.,
- Retrieval Mechanisms: BM25, Dense Passage Retrieval (DPR), Hybrid Search etc.,
Machine Learning & Deep Learning
- Algorithms: XGBoost, LSTM, CNN, Transformers, Random Forests etc.,
- Techniques: Transfer learning, Hyperparameter tuning, model training, and deployment.
Programming, Databases, and Frameworks
- Languages: Python
- ML/DL Frameworks: PyTorch, TensorFlow
- Libraries: NumPy, Pandas, Scikit-learn
- API Development: FastAPI, Flask
- Databases: SQL, NoSQL (MongoDB, Cassandra)
Cloud & Deployment
- Cloud Platforms: Azure Machine Learning, Azure Functions, Azure Databricks.
- Containerization: Docker
Required Skills
- Expertise with Pretrained LLMs, fine-tuning, RAGs, and deployment.
- Proficiency in Azure cloud environment, including model hosting and scaling.
- Advanced skills in Python, ML/DL frameworks
- Strong problem-solving, data analysis, and domain modelling capabilities.
- Demonstrated ability to mentor junior team members effectively.
Preferred Skills
- Experience with knowledge graphs, semantic search, and conversational AI.
- Familiarity with edge AI and real-time model inferencing.
- Knowledge of domain-driven design and advanced deployment strategies.
- Experience in managed service.
EY exists to build a better working world, helping to create long-term value for clients, people and society and build trust in the capital markets.