Technology

A dedicated team of Data/AI Scientists, Data/AI Engineers, MLOps/LLMOps Engineers, Operations Research Scientists, Supply Chain Management Specialists, Research Scientists, Business Strategists and Business Analysts under the leadership of our Chairman and Chief Scientist Mr. Sougata Biswas are ready to understand your business problem and resolve.

Tools sets:

  • ML Tools: R, Python
  • DL Tools:Tensorflow, Keras, Pytorch
  • Version Control: GIT
  • Cloud: AWS, Azure, GCP
  • Project Management: Agile (Scrum)
  • GraphDB: Neo4j
  • MLOps : MLFlow, Kubeflow
  • Generative AI: Open AI, HuggingFace, LangChain, Databricks
  • AR-VR, Drone; Quantum Computing, GEO (Generator Enhanced Optimizer

Core Competency:

  • Provide Enterprise level Traditional & Cognitive product development using traditional ML techniques as well as advanced state of the art Deep Learning
  • Subject Matter Expert in NLP/NLG, Speech/Sound, Computer Vision (Image/Video, GAN), QnA / Conversational Virtual Assistants, Prediction/Forecasting, Time Series Analysis, Social Media Analytics, Semantic Knowledge Graph etc.
  • Expert in 3 branches of Generative AI: Large Lange Models(LLMs), Large Vision Models (LVMs) & Generator Enhanced Optimization (GEO).
  • Optimization Techniques; Different techniques for Operations research & Game Theory
  • Extensive experience in driving Data Science practice; analyzed problems and uncovered opportunities using predictive analysis. Achievement oriented professional with excellent people management skills and an ability to manage change with ease.
  • Cloud Computing (Azure, AWS & GCP) expert & Azure Certified Professional to provide Enterprise level E2E Product development.
  • Proficient in Atlassian JIRA & Confluence and managing a project with Agile/Scrum project management principles.
  • Prompt Engineering for LLMs using HuggingFace, Open AI, LangChain, LaamaIndex, Llama 2/3, Mistral AI
  • Computer Vision; Generative Adversarial Networks (GANs); Prompt Engineering for Large Vision Models using Comet; Deep Fake Images / Videos Generation
  • Fine-tuning of LLM with instruction prompts, Parameter Efficient Fine-tuning (PEFT)[LoRA, QLoRA, Adapters, Prompt-tuning with Soft Prompt) and Quantization Techniques using Hugging Face
  • Fine-tuning of LLMs Reinforcement Learning from Human Feedback (RLHF)
  • LLMOps using MLFlow / AirFlow / Vertex AI / Kubeflow Pipeline
  • Evaluation LLMs using LlamaaIndex & TruLens; Different Advanced RAG Techniques