GenAI Data Scientist

Job Description

- Model Development: Research, design, and develop state-of-the-art generative models such as GPT, GANs, VAEs, or diffusion models for tasks like text generation, summarization, reasoning, Q&A, or predictive analytics.
- Fine-Tuning: Fine-tune pre-trained models (e.g., Llama3, Mistral, OpenAI GPT, BERT, T5) to meet domain-specific requirements.
- Data Preparation: Clean, preprocess, and structure large datasets to train, validate, and test generative AI models.
- Deployment: Implement scalable solutions for deploying generative AI models in production environments using tools like Docker, Kubernetes, or cloud platforms (AWS, GCP, Azure).
- Collaboration: Work closely with cross-functional teams, including product managers, engineers, and stakeholders, to align AI solutions with business goals.
- Evaluation: Develop metrics and benchmarks to evaluate model performance and ensure quality outputs.
- Research: Stay up to date with the latest advancements in AI/ML, particularly in generative AI techniques, and propose innovative solutions.
- Ethical AI Practices: Ensure ethical considerations are addressed in model development, including fairness, accountability, and explainability. Experience of architecting AI systems to solve complex business problems.
- Build advanced RAG pipelines, text chunking, and retrieval, LLM Prompt Engineering, using Vector Databases. Implement right LLM selection based on use cases and client criteria (GPT-4, Llama2, Mistral, Claude, Gemini, Flan, BERT) while managing trifecta of accuracy, cost, and latency/scale.
- Develop, fine-tune, context tune and implement state-of-the-art NLP models including Large Language Models like GPT, Llama3.1, Claude, BLOOM, Flan-T5, Falcon etc. fine-tuning PEFT LoRa/QLoRa adapters
- design AI systems & architectures, while considering good Responsible AI standards and AI Governance.
- Build Agentic AI, AutoGen, Muti-agents use cases, AI Autonomous Agents and LLM orchestration architectures for enterprise data at scale in Python.
- Hands on experience with complementary technologies around LLMs like embedders, vector databases (chroma, weaviate etc.), orchestration tools like Langchain, LlamaIndex etc.
- Conduct research and experimentation to improve existing models and propose novel approaches.
- Collaborate with cross-functional teams to integrate generative AI solutions into real-world applications.
- Stay up-to-date with the latest advancements in deep learning and generative models and apply them to enhance our AI capabilities.
- Document research findings, prepare technical reports, and contribute to whitepaper/scientific publications.
- Provide deep leadership and coaching in the project delivery lifecycle. Focus on shared learning, continuous improvement, and drive adoption of best practices.

Qualifications

Education and Experience:

- Masters of Science or PhD in computer science, data science, statistics, Natural Language Processing
- 2-3 years of experience in GenAI and Large Language Models.
- Strong programming skills in Python, including experience with libraries like TensorFlow, PyTorch, Hugging Face Transformers, or similar. as evidenced by released code (e.g., GitHub repositories version control awareness).
- Solid understanding of optimization techniques for training deep neural networks, regularization methods, and hyperparameter/fine tuning.
- Hands-on experience with generative AI models, such as Llama3, GPT4, Claude
- Knowledge of NLP, computer vision, or multimodal AI techniques.
- Proficiency in data manipulation and analysis tools (e.g., Pandas, NumPy, SQL).
- Experience in Generative AI Models and LLMs, finetuning LLMs, prompt engineering and experience with LLM orchestration frameworks like Langchain, LlamaIndex, RAGAS, etc.
- Strong software engineering skills for rapid and accurate development of AI models and systems.
- Understanding of Agentic AI, Autonomous Agents, AI Agents, AutoGen, Crew.ai, Langchain, and workflow steps design. NVIDIA Blueprints. Google AI Agents.
- Provide business-oriented solution with ability to communicate effectively, both verbally and in writing, with technical and non-technical stakeholders.
- Experience working in a collaborative environment, contributing to multidisciplinary teams and projects.
- Proven ability to solve complex problems, think creatively, and adapt to evolving research trends.
- Experience in deploying ML models in cloud environments (AWS SageMaker, GCP AI Platform, or Azure ML).
- Strong problem-solving and analytical skills.
- Excellent communication skills, both technical and non-technical.

Skills and Competencies:

- 5 years of experience in AI, NLP including transformer architecture and LLMs, Computer Vision and related technologies
- Excellent communication and problem-solving skills.
- Working with cross-functional teams across different stakeholders.
- Ability to explain GenAI to non-technical audiences across many different industries. Ability to do GenAI Architecture solutioning for existing clients and potential clients.
- People leadership skills. Candidate is also hands on developer while people managing Data Scientists, ML Engineers etc.
- Experience in ML Engineering and MLOps, MLFlow
- Strong understanding of statistical and machine learning concepts
- Experience with deep learning frameworks such as TensorFlow and PyTorch
- Familiarity with key concepts and techniques used in generative models, such as variational autoencoders (VAEs), generative adversarial networks (GANs), and flow-based models.
- Strong programming skills in languages such as Python, along with experience working with popular deep learning frameworks like PyTorch and TensorFlow.
- Understanding of Graph Database and/or Vector Database along with knowledge of cloud services (e.g., AWS, Azure, GCP).
- Experience with deploying AI models in production environments.
- Familiarity with domain-specific applications of generative AI
- Leveraged both Azure and AWS for model inferencing. For finetuning, he has worked more on AWS SageMaker.
- Has used different retrieval, reranking, and generation techniques based on the use case complexity and the metrics required.
- Has played multiple roles in the past, such as an individual contributor, a solution architect, a mentor, a trainer, and a people manager.
- He also participates in RFI, RFP, thought leadership, and business development activities

---