Responsibilities

- Design, build, and deploy A2A communication architectures and multi-agent workflows for autonomous collaboration, delegation, and multi-step business processes.
- Integrate proprietary and open-source LLMs with enterprise APIs, tool-calling methods, and third-party SaaS applications.
- Implement RAG, enterprise grounding, vector search, and hybrid search strategies for reliable inter-agent operations.
- Provision and manage AWS, Azure, and GCP environments using Terraform for high-concurrency LLM inference and agent coordination.
- Create CI/CD pipelines and support testing, deployment, versioning, LLM evaluation, and agentic workflow lifecycle management.
- Apply security and network controls including VPCs, secure API gateways, encryption, IAM, and protected inter-agent communication.
- Build observability and evaluation systems to monitor agent interactions, accuracy, token spending, model drift, and agent loops.
- Document LLM integration protocols, A2A interaction flows, and cloud infrastructure deployments.

Requirements

- Deep experience integrating, fine-tuning, and optimizing foundation models, including prompt engineering, tool calling, and structured outputs.
- Proven experience building multi-agent systems, inter-agent messaging pipelines, state management frameworks, and task delegation protocols.
- Hands-on experience with LangGraph,



LangChain, AutoGen, LangSmith, or comparable agentic and tracing platforms.
- Strong understanding of AI interoperability standards including MCP and open multi-agent communication specifications.
- Advanced proficiency in Python or TypeScript for backend orchestration, agent memory systems, and API service development.
- Practical knowledge of RAG, vector databases, hybrid search architectures, and context management.
- Hands-on experience with DevOps and infrastructure automation using Terraform, Docker, and Kubernetes.
- Solid knowledge of secure cloud architecture, zero-trust networking, private endpoints, OAuth, SAML, and IAM.
- Experience with logging, distributed tracing, metrics, and evaluation of non-deterministic multi-agent workflows.
- Knowledge of enterprise agentic and workflow platforms such as Workday A2A, Salesforce Agentforce, and ServiceNow AI Agents.

Benefits

- Distributed-company work model with flexible locations and schedules for many roles.
- Health coverage for employees and families in many locations.
- Generous vacation allowance.
- Up to $2,000 matching for financial donations and up to 40 volunteer hours annually.
- At least 16 weeks of parental leave.
- Inclusive culture and equal employment opportunity.

📌 Agentic AI Engineer (Bengaluru)
🏢 Elastic
📍 Bengaluru