Primary skills : Gemini Agent space, AI multi cloud experience, Experience in running AI services, vibe coding(Cursor), Vertex AI, Vector DB, RAG Observability, LLM/SLM Evaluation metrics, Vector Search Performance Evaluation, Extensive hands-on with Python and LLM application frameworks (like LangChain, LangGraph etc)

Secondary skills : Terraform, Multi cloud experience (AWS), Model Fine-Tuning, ETL, PySpark, Java

Updated JD

JD below.

• Extensive implementation experience in data analytics space or a senior developer role in one of the modern technology stacks.  
• Experience with Data Science techniques. Knowledge of data warehousing concepts, including data warehouse technical architectures, infrastructure components, ETL/ELT and reporting/analytic tools and environments. 
• Excellent programming skills and proficiency in at least one of the major programming scripting languages used in Gen AI orchestration such as Python and PySpark/Java  
• Ability to build API based scalable solutions and debug & troubleshoot software or design issues. Strong understanding of API gateways load balancing and rate limiting strategies.
• Hands-on experience implementing AI safety and guardrails across the full pipeline including input validation, PII detection, reliability checks and compliance
• Knowledge on multi-layered security (access control, data isolation, PII redaction, audit logging)
• Knowledge on metric framework like RAGAS and RAG metrics (Precision@K, Recall@K, MRR, nDCG, Faithfulness, Answer Relevance) and design/operating RAG at scale
• Knowledge on concrete privacy, safety, and bias mitigation techniques in enterprise settings
• Deep understanding of vector search and indexing concepts
• Hands on exposure to integrating with models like Gemini Pro, GPT etc., using API endpoints.  
• Thorough understanding of prompt engineering; implementation exposure to LLM application frameworks like LangChain & vector databases and agentic orchestration frameworks like LangGraph
• High level knowledge on fine tuning pretrained models using techniques like SFT, Domain specific fine-tuning over custom datasets
• Ability to quickly conduct experiments and analyze the features and capabilities of newer versions of the LLM models as they come into market