Data Scientist II

Location: Bengaluru, India

Department: Projects & Delivery

Skills: Data Science, Machine Learning, Agentic AI, generative AI

• GenAI & Agents: Build and deploy Generative AI solutions — agentic systems and RAG pipelines — including agents that reason, use tools, and act over structured (databases, APIs) and unstructured data.
• Multi-Agent Orchestration: Design multi-agent workflows using patterns like ReACT, planner–executor, and agent-critique to break down and solve complex tasks.
• Reliability & Evaluation: Establish evaluation frameworks — LLM-as-a-judge, groundedness/faithfulness, hallucination detection — to test and ensure the safety and accuracy of LLM systems.
• End-to-End ML: Apply forecasting, CLV, recommendation, and NLP techniques, owning the full lifecycle from prototype to scalable, low-latency deployment (Docker, AWS).
• Big Data: Process billions of records with PySpark to engineer features and extract actionable insights.
• Experimentation: Run experiments and EDA to validate hypotheses, and communicate results and solution outlines clearly to stakeholders.
• Mentorship: Mentor junior team members and bridge business problems and data science across engineering and product teams.

• 3–4 years building and deploying machine-learning models in production.
• Generative AI with LLMs — RAG (chunking, embeddings, vector search with Pinecone/Milvus/Weaviate, re-ranking) and agentic frameworks (LangChain, LangGraph).
• Hands-on with tool use / function calling, prompt engineering, and structured outputs / guardrails.
• Strong classical ML and deep learning (ANN, CNNs, LSTMs, Transformers).
• Traditional NLP: Transformers (BERT, T5, GPT), Word2Vec, NER, topic modeling, contrastive learning.
• Expert Python (Pandas, NumPy, Scikit-learn, PyTorch/TensorFlow) and strong SQL; PySpark for distributed processing.
• Solid system design and MLOps — Docker and AWS (S3, Lambda, ECR, Step Functions).

• Pharma or life-sciences domain, with claims data and commercial analytics.
• Fine-tuning open-source LLMs (Llama, Mistral) for specific tasks.
• Latest agentic systems — MCP and multi-agent frameworks (AutoGen, CrewAI).
• Advanced ML: PU learning, representation learning, and model explainability.
• Domain models such as Marketing Mix Models (MMM), demand forecasting, or multi-dimensional time-series.
• Open-source AI/ML contributions or publications.

• Competitive ESOP grants.
• Working with Fortune 500 companies and world-class teams.
• Publishing papers and attending conferences.
• Networking events, conferences, and seminars.
• Visibility across all functions at Akaike — sales, pre-sales, lead generation, marketing, and hiring.