Data Scientist - job 3 of 4

Dice is the leading career destination for tech experts at every stage of their careers. Our client, Cardinal Integrated Technologies Inc, is seeking the following. Apply via Dice today!Role Name: Data Scientist with Gen AIDuration: 6-12+ MonthsLocation: Remote Eastern Time ZoneSkill Summary: Generative AI - Data Scientist SMEProject Staff Augmentation requirements for a Data Scientist Consultant aligned to AI EngagementsOpportunity OverviewThe client is looking for a technical data scientist focused on designing and developing AI models. Specializing in training, testing and tuning Generative AI models, leveraging data (pipelines/assets/APIs) provided by the data engineer.As a consultant within the organization, aligned to consulting service delivery, you will provide consultative business and technical delivery services for a range of complex customer-facing consulting engagements. Specifically, you will be part of projects deploying, administering, and configuring Large Language Model, Generative and other AI software stacks and their associated models. These projects will focus on helping Dell customers to implement AI Use cases for their business.Essential Requirements And Skills• 10+ years working with customers in the Data Science and Data Analytics field.• 5+ years developing AI Solutions for large customers.• 2+ years working in AI Solutions that leverage Neural Network based solutions, especially within the context of NLP.• Extensive knowledge of statistical analysis, modelling, transformer architectures and word embeddings.• Very strong knowledge of LLMs, prompt engineering, guardrails, model chaining, fine tuning, and training. Experience with supervised and unsupervised training, PEFT, LoRA and RLHF.• Detailed knowledge of latest developments in LLM related techniques such as external data integration using RAG, Reinforcement learning, Transfer Learning and Data Annotation. Experience with related tools such as Nvidia RAG LLM Playground, Nvidia Query Router, Nvidia Guardrails and Nvidia AI Workbench• Detailed knowledge of embeddings and embeddings models. Ability to evaluate embeddings engines and perform pre- and post-retrieval optimization.• Ability to leverage NoSQL (such as Redis) and Vector Databases. Good SQL knowledge is required.• Experience with working with Vector Databases such as PGVector and Milvus.• Extensive experience and knowledge with using Python, PyTorch, TensorFlow, TensorRT, and Jupyter notebooks.• Good knowledge of Gradio.• Knowledge of CUDA libraries, Langchain modules, LlamaIndex, Pandas and other related libraries.• Knowledge of Mistral LLM and Gemma LLM Architectures.• Hands on knowledge of the Nvidia AI tool suite such as NeMo Framework, Triton Inference Server, Nvidia NIMs, Nemo Retrieved Embeddings Microserver (NREM), Nvidia ChainServer and others.• Hands on knowledge and experience of working on AMD GPUs and the associated AMD AI tool suites such as ROCM, OpenMP, OpenCL, HiPify and MIOpen.• Good working knowledge of Hugging Face Portal and Interfaces, including usage of Hugging Face Models.• Experience with utilizing LLMs to support the application development lifecycle (SDLC), such as using it for code completion, test analysis and code summarization. Experience with defining an ROI strategy when introducing AI in the SDLC.• Familiarity with Linux.• Experience using popular MLOps and AIOps tools such as Cvrg.io, Kubeflow, mlflow, Kserve, vLLM and Torchrun.• Experience with using techniques to optimize large models.• Ability to evaluate models using well known metrics such as BLEU and Rouge. Ability to consult customers on the best models aligned to specific customer use cases using validation metrics, probability distributions, statistical inference and hypotheses testing. Ability to define and implement performance metrics to evaluate scalability of AI solutions.• Ability to built test frameworks for AI models.• Ability to consume API endpoints for data querying and ingestion• Ability to gain customer trust, ability to plan, organize and drive customer workshops. Good communication skills in English is a must.• The ability to work in a highly efficient team using an Agile methodology such as Scrum or Kanban. Ability to have extended pairing sessions with customers, enabling knowledge transfers in complex domains.• Ability to influence and interact with confidence and credibility at all levels within the Dell Technologies companies and with our customers, partners, and vendors.• Experience working on project teams within a defined methodology while adhering to margin, planning and SOW requirements.• Ability to be onsite during customer workshops and enablement sessions.Desirable Requirements And Skills• General awareness of Dell Technologies products.• Knowledge and understanding of IBM Watson ecosystem.• General knowledge of TensorBoard, Hugging Face AutoTrain and Hugging Face Text Generation Inference.• Operational Knowledge of Vendor supported Kubernetes distributions such as OpenShift, Tanzu and Rancher.• In possession of the Certified Kubernetes Developer (CKD) certification is a plus.• University Degree aligned to Machine Learning or Data Science is a plus.--Best Regards:Asif Khan |Assistant Manager Recruitment | Cardinal Integrated Technologies