Lead Systems Engineer - Google Forward Deployed Engineer

EPAM Systems; India; On-site; INR 2,500,000 - 6,000,000; Full time

Job summary

EPAM Systems Inc in India seeks a senior AI/ML engineer to design and deploy production-grade multi-agent systems using Gemini Enterprise, with deep integration into Google Workspace, NotebookLM, and Vertex AI. You will implement secure authentication, data governance, and scalable state management across enterprise environments.

You will prototype with Agent CLI, apply advanced reasoning patterns like CoT and ReAct, and drive CI/CD for agent deployments. 7–13 years IT experience preferred.

Qualifications

• 7–13 years of IT experience with 4–6+ years in software development, cloud solution architecture, and AI engineering, preferably on GCP.
• Proficiency in Python with production-grade software engineering practices, and building multi-agent systems using Google ADK or comparable frameworks.
• Experience deploying Gemini Enterprise, NotebookLM, and integrating with Google Workspace APIs, Add-ons, and Drive/Docs connectors.
• Knowledge of authentication/authorization protocols across GCP and 3rd-party SaaS platforms (IAM, OAuth 2.0, OIDC).
• Proven track record of designing, building, and deploying production AI agents or GenAI solutions using ADK, Vertex AI, or Gemini.

Responsibilities

• Design, develop, and deploy production-grade Multi-Agent Systems and agentic workflows
• Identify need for Pro Code vs No Code design based on customer problem statements
• Build and deploy custom agents using the ADK and integrate with Gemini Enterprise
• Architect scalable session storage, memory architectures, and state persistence for multi-agent interactions
• Implement end-to-end authentication and authorization perimeters for agents and integrations
• Enforce data governance policies and enterprise ACLs across data sources
• Utilize Agent CLI and Antigravity for local testing, tracing, and RCA
• Architect HITL workflows for Interrupt and Resume patterns between agents and humans
• Build automated CI/CD pipelines for agent deployments, testing, and regression evaluations
• Incorporate CoT, ReAct, Plan-and-Solve, and Self-Reflection reasoning patterns
• Collaborate with client engineering teams to deploy enterprise AI solutions in client environments

Skills

Python

Multi-Agent Systems

GCP

IAM/OIDC

CI/CD

Prompt engineering

Tools

Gemini Enterprise

NotebookLM

Google Workspace

Vertex AI

Agent CLI

MCP Connectors

Job description

Responsibilities

• Design, develop, and deploy production-grade Multi-Agent Systems and agentic workflows using Gemini Enterprise
• Identify the need for Pro Code vs No Code design based on customer's problem statement and data sources
• Build and deploy custom agents using the ADK and Agent Platform AI Reasoning Engine, and integrate them with Gemini Enterprise
• Build native integrations across the Google Workspace Ecosystem and NotebookLM for grounded enterprise research, document automation, and conversational discovery
• Architect scalable short-term session storage, conversation state persistence, and long-term memory architectures across multi-agent interactions
• Implement end-to-end authentication and authorization perimeters for agents, GCP services, and 3rd-party SaaS integrations using IAM, OAuth 2.0, OIDC, and service identities
• Enforce fine-grained ACLs, user/group permission inheritance, and data governance policies to ensure agents strictly adhere to enterprise access boundaries
• Utilize developer tooling including Agent CLI and Antigravity for local rapid prototyping, testing, emulation, distributed tracing, and RCA
• Architect HITL workflows, ensuring seamless Interrupt and Resume patterns between autonomous agents and human experts for critical decision points
• Build and maintain automated CI/CD pipelines to manage versioned agent deployments, automated testing, and continuous regression evaluations
• Implement advanced reasoning patterns such as Chain-of-Thought (CoT), ReAct, Plan-and-Solve, and Self-Reflection to enhance agent reliability
• Collaborate directly with client engineering and architecture teams to deploy, validate, and operationalize enterprise agentic AI solutions in client environments

Requirements

Requirements 7-13 years of overall IT experience, including 4 to 6+ years in software development, cloud solution architecture, and AI engineering, preferably on GCP

• Proficiency in Python with production-level software engineering practices, and expertise in building multi-agent systems using Google ADK or comparable agentic frameworks
• Hands-on experience deploying and customizing Gemini Enterprise, NotebookLM, and integrating with the Google Workspace Ecosystem, including APIs, Add-ons, and Drive/Docs connectors
• Expertise in managing agent session state, short-term context caching, and durable long-term memory backends such as vector databases and relational/NoSQL session stores
• Experience with MCP Connectors to securely integrate core third-party platforms with Gemini Enterprise
• Understanding of authentication and authorization protocols across GCP and 3rd-party SaaS platforms, including IAM, OAuth 2.0, and OIDC
• Knowledge of enterprise ACLs, identity federation, permission trimming, and RBAC across heterogeneous data sources
• Proficiency in developer tooling such as Agent CLI and Antigravity for local simulation, debugging, prompt engineering, and execution graph tracing
• Proven track record of configuring and maintaining automated CI/CD pipelines for agent containerization, deployment, and test automation
• Knowledge of AI evaluation methodologies, trajectory analysis, and performing RCA on agent execution failures, tool errors, and prompt regressions
• Proven track record of designing, building, and deploying production AI agents or enterprise GenAI solutions using ADK, Vertex AI, or Gemini
• Hands-on experience in customer-facing technical delivery, solutions engineering, or FDE roles, with familiarity with enterprise Agile SDLC, sprint delivery, and stakeholder management
• Nice to have Experience with inter-agent communication protocols such as MCP and A2A architectures
• Experience integrating agents with Enterprise Knowledge Graphs, structured ontologies, and hybrid vector search systems
• Experience with enterprise data security frameworks such as Sensitive Data Protection and VPC Service Controls