Junior Data Scientist

Design, train, and deploy Machine Learning and Deep Learning models for structured and unstructured data across domains (text, image, and audio).

• Work on Agentic AI pipelines that combine reasoning, planning, and multi-tool orchestration using frameworks like LangGraph, LangFlow, n8n, and Flowise.
• Develop Bedrock-powered Agents leveraging Claude 3, Mistral Large, or Llama models for enterprise-grade automation and decision intelligence.
• Implement Retrieval-Augmented Generation (RAG) and Context-Aware Generation (CAG) frameworks for intelligent document understanding, conversational search, and contextual QA.
• Build Conversational AI bots that integrate with AWS Bedrock, SageMaker, Lex, and OpenSearch to deliver natural, human-like interactions.
• Apply Prompt Engineering and Prompt Chaining techniques for reasoning-based LLM workflows, contextual memory, and agent communication.
• Develop and fine-tune state-of-the-art NLP models (BERT, GPT, Claude, LLaMA, T5, etc.) for text summarization, classification, and semantic retrieval.

Develop Conversational AI chatbots and copilots integrated with Bedrock, SageMaker, Lex, and OpenSearch, enabling intelligent query handling and reasoning-based responses.

• Design and train Computer Vision models for image recognition, OCR/ICR, and visual data analytics using frameworks like OpenCV, Detectron2, YOLO, and Amazon Rekognition.
• Architect scalable multi-agent systems that perform task decomposition, self-validation, and reinforcement-based optimization.
• Integrate external APIs and enterprise systems (SAP, CRM, ERPs, etc.) within agentic workflows for end-to-end process automation.
• Work with vector databases (OpenSearch, Pinecone, FAISS, or Chroma DB) for knowledge grounding and enterprise data retrieval.

Manage cloud infrastructure deployments using AWS EC2, Lambda, API Gateway, Step Functions, and CloudFormation.

Implement CI/CD pipelines and MLOps best practices for automating model training, validation, and deployment.

Collaborate with data engineers and DevOps teams to ensure reproducibility, observability, and performance tuning of AI workloads.

Apply Prompt Engineering, Prompt Chaining, and Tool Use patterns for reasoning-enhanced LLM-based applications.

Monitor production systems and implement model drift detection, retraining triggers, and performance alerts.

• Contribute to building Generative AI solutions text-to-SQL, text-to-image, text-to-speech, and autonomous workflow assistants.

• Evaluate and implement state-of-the-art foundation models (e.g., Claude 3.5, Gemini 2.0, Mistral Large, Mixtral, Titan) for specific domain use cases.
• Collaborate closely with data engineers and cloud architects to deploy scalable solutions on AWS (Bedrock, SageMaker, Lambda, API Gateway, ECS/EKS, Q)  working on multi cloud platforms like Azure or GCP will be plus.
• Document ML pipelines, maintain experimentation logs, and ensure reproducibility through MLOps best practices.
• Participate in research, benchmarking, and continuous learning to stay ahead of the curve in the fast-evolving Generative AI and Agentic AI landscape.

Requirements

Qualifications:

• B.Tech / M.Tech / MS in Computer Science, Artificial Intelligence, Data Science, or related fields.

• Strong mathematical foundation in Statistics, Probability, and Linear Algebra.
• Coursework or AWS certifications in Machine Learning, Deep Learning, or Generative AI are a plus.

Key Skills & Attributes:

• Programming Proficiency: Strong experience with Python, and libraries such as Pandas, NumPy, Scikit-learn, TensorFlow, PyTorch, Hugging Face Transformers, and LangChain.

• Agentic AI Tools: Hands-on exposure to LangGraph, LangFlow, Flowise, and n8n for building multi-agent pipelines and workflow automation.
• Generative AI Expertise: Understanding of Bedrock LLMs (Claude, GPT, Llama, Mistral, Gemini) and their applications in reasoning, summarization, document intelligence, and conversational AI, sagemaker AI,  Lambda, API Gateway, Comprehend, Textract, and OpenSearch for developing and deploying AI workloads..
• Prompt Engineering: Experience with prompt design, chaining, retrieval context optimization, and few-shot learning techniques.
• RAG / CAG Systems: Knowledge of Retrieval-Augmented Generation (RAG) and Context-Aware Generation (CAG) frameworks for contextual grounding and adaptive reasoning.
• MLOps & Cloud Deployments: Practical understanding of Docker, Kubernetes (EKS/ECS), EC2, Lambda, Step Functions, and CloudFormation for scalable deployments.
• SQL/NoSQL/Vector Databases: Experience with Pinecone, FAISS, or OpenSearch or Pgsql or MongoDB Vector Engine for semantic search and embedding management.
• Model Deployment & Integration: Ability to design APIs for serving ML models using FastAPI, Flask, or AWS API Gateway.
• Data Handling: Strong skills in data preprocessing, feature engineering, data wrangling, and EDA for ML readiness.
• NLP & Computer Vision: Exposure to transformer-based NLP models and CNN-based CV architectures for text and image understanding.
• DevOps Collaboration: Understanding of CI/CD pipelines, Git-based workflows, and container orchestration for continuous delivery.
• Analytical & Mathematical Foundations: Solid grounding in Probability, Statistics, Linear Algebra, and Optimization for model building.
• Research Orientation: Eagerness to explore SOTA models (e.g., Claude 3.5, Gemini 2.0, Mistral Large) and experiment with multi-agent orchestration and reasoning frameworks.
• Soft Skills: Excellent problem-solving attitude, curiosity-driven mindset, and collaborative approach towards multi-functional teamwork.
• Continuous Learning: Passion for staying updated with AI/ML innovations, Agentic AI frameworks, GenAI APIs, and Cloud-native ML advancements.