Offensive Security Mts/smts - AI Platform & Orchestration - Hyderabad

Salesforce Salesforce · Enterprise · Hyderabad, India

This role is for a backend-focused engineer to build an AI platform for offensive security operations. The focus is on creating the infrastructure, AI harness, and orchestration layers for autonomous security workflows, including agent-based automation, state management, memory, and tool integration for LLM-based systems.

What you'd actually do

  1. Build and maintain agent-based AI automation workflows and the core components of our security platform, including state management, memory strategies, and context handling.
  2. Develop scalable backend services and APIs that integrate with existing internal & external data sources to gather signals on potential targets.
  3. Implement and optimize tool interfaces and execution logic for LLM-based systems, ensuring predictable and observable behavior.
  4. Work closely with security practitioners and researchers to ensure platform outputs are actionable, relevant, and grounded in real-world data.

Skills

Required

  • Python, Go, or Java
  • API integration
  • distributed services
  • large-scale data ingestion
  • backend development
  • platform development
  • system design

Nice to have

  • RAG
  • MCP (Model Context Protocol) servers
  • Agent-to-Agent (A2A) orchestration frameworks
  • workflow engines
  • signal correlation systems
  • autonomous DAST/SAST tooling
  • operationalizing LLMs beyond simple prompt-response
  • stateful memory
  • task decomposition
  • offensive security concepts
  • red teaming
  • vulnerability research

What the JD emphasized

  • backend-first engineer
  • modular, extensible systems
  • reason over large data sources
  • execute multi-step workflows
  • agent-based AI automation workflows
  • state management, memory strategies, and context handling
  • tool interfaces and execution logic for LLM-based systems
  • predictable and observable behavior

Other signals

  • building AI platform
  • autonomous offensive security operations
  • agent-based AI automation workflows
  • orchestration layers