7 -10 years
- More on DevOps, and AI Ops (used when Agentic AI comes into picture)
Platform & Framework Development
- Build reusable agentic AI frameworks, orchestration templates, and accelerators
- Develop shared libraries for:
- Tool orchestration
- Agent communication
- Memory handling
- Workflow management
- Evaluation pipelines
- Create standardized development patterns for enterprise AI systems
MCP & Integration Enablement
- Develop and maintain MCP-compatible integrations and enterprise connectors
- Build reusable APIs and integration services for enterprise platforms
- Enable scalable access to enterprise data sources and business tools
Governance & TRiSM Enablement
- Define lightweight governance standards for agentic AI systems
- Implement traceability, monitoring, logging, and lifecycle tracking mechanisms
- Support trust, risk, security, and monitoring (TRiSM) compliance
- Establish evaluation, observability, and auditability practices for AI agents
Engineering Enablement
- Create reusable templates, starter kits, and deployment accelerators
- Improve developer productivity through standardized tooling and automation
- Establish documentation standards and reusable implementation guides
Operational Excellence
- Support production readiness for AI systems
- Define monitoring, telemetry, guardrails, and fallback strategies
- Collaborate with solution teams to improve reliability and maintainability
Required Skills
Technical Skills
- Strong software architecture and platform engineering experience
- Expertise in:
- LangGraph / Semantic Kernel / AutoGen / CrewAI
- Orchestration frameworks
- AI system architecture
- MCP ecosystem concepts
- Strong backend engineering experience using Python/Node.js
- Experience building reusable platforms/frameworks
- Robust understanding of:
- AI governance
- Security
- Observability
- Evaluation frameworks
- Monitoring systems
- Experience with APIs, event-driven architectures, and cloud-native systems
Good to have
📌 Data Software Engineer-AI (Pune)
🏢 Cummins
📍 Pune