Agentic AI

Netskope

Bengaluru 15 Years Exp Posted 202d ago

15+ years of software engineering experience with demonstrated progression to senior technical roles
Deep expertise in distributed systems architecture at massive scale (billions of events/day, petabyte-scale data)
Proven track record architecting AI/ML systems in production environments, including experience with:
- Large Language Models (LLMs) and generative AI applications
- Retrieval-Augmented Generation (RAG) architectures and vector search systems
- Machine learning model deployment, monitoring, and MLOps practices
- Real-time inference systems and online learning
Expert knowledge of graph databases and graph algorithms (Neo4j, TigerGraph, or similar) including:
- Graph data modeling and schema design
- Complex graph queries and traversals (Cypher, Gremlin, or similar)
- Graph algorithms for community detection, centrality, path finding
- Large-scale graph processing and analytics
Master-level proficiency in Python and modern ML/AI frameworks (TensorFlow, PyTorch, LangChain, LangGraph)
Deep understanding of data engineering including:
- Streaming data architectures (Kafka, Flink, Pulsar)
- Large-scale data storage (ClickHouse, Snowflake, BigQuery, data lakes)
- ETL/ELT pipeline design and optimization
- Real-time and batch processing paradigms

Hands-on experience building AI-powered security solutions such as:
- Behavioral analytics and user/entity behavior analytics (UEBA)
- Anomaly detection using unsupervised and semi-supervised learning
- Threat classification and automated triage systems
- LLM-based security assistants and conversational interfaces
- Graph neural networks for security relationship modeling
Expertise in vector databases (Pinecone, Weaviate, Chroma, Milvus, pgvector) and their application to:
- Semantic search over security data
- Threat intelligence matching and similarity analysis
- Security knowledge base construction
Deep understanding of embedding models and semantic representation of security concepts
- Experience with prompt engineering, fine-tuning, and LLMOps best practices