Senior Agentic AI Engineer – Google ADK / Gemini / GCP

Infosys Pontoon•Richardson, TX

3d•Hybrid

About The Position

We are seeking a skilled Senior Engineer - Agentic AI with hands-on expertise in Google Agent Development Kit (ADK), Gemini models, and Google Cloud Platform. The successful candidate will build reliable agentic solutions that translate business requirements into scalable, secure, observable, and maintainable production systems. The role combines strong software engineering, agent architecture, evaluation, cloud deployment, and technical leadership. The engineer will own the end-to-end agent lifecycle, from use-case discovery and design through production operations, feedback-driven enhancement, and governance.

Requirements

Agentic AI Google ADK, Gemini & Google Cloud
Good Communication Skill
Good Technical Skill
Minimum 5-8 years of experience

Nice To Haves

RAG architectures, semantic search, embeddings, vector databases, Vertex AI Vector Search, and knowledge graphs.
Model Context Protocol (ClientP) and secure tool or connector integration patterns.
Experience with complementary agent frameworks such as LangGraph, LangChain, CrewAI, AutoGen, or similar technologies.
Human-in-the-loop and approval-based workflows for high-impact actions.
Terraform or equivalent infrastructure-as-code tooling.
Responsible AI, privacy, threat modeling, prompt-injection defense, content safety, and AI governance.
Integration experience with enterprise platforms such as CRM, ERP, ITSM, collaboration, workflow, and data platforms.

Responsibilities

Design, develop, test, and deploy AI agents using Google ADK.
Build single-agent and multi-agent systems for reasoning, planning, task execution, collaboration, and delegation.
Develop custom tools, function-calling interfaces, workflow agents, and integrations with enterprise APIs, databases, SaaS platforms, and internal systems.
Implement memory, session and state management, context management, retrieval-augmented generation (RAG), and knowledge-grounding patterns.
Apply deterministic workflows and dynamic agent orchestration patterns based on solution needs.
Own the agent lifecycle from ideation, prototyping, and validation through release, production operation, retirement, and replacement.
Define versioning, configuration, release, rollback, and environment-promotion practices for agents, prompts, tools, policies, and models.
Establish observability using structured logs, traces, metrics, execution trajectories, and error analysis.
Implement security, access control, data protection, governance, and responsible AI controls.
Improve reliability, scalability, resilience, latency, throughput, and cost efficiency in production.
Design automated and human-in-the-loop evaluation frameworks covering task success, accuracy, groundedness, response quality, safety, tool-use effectiveness, latency, and cost.
Create benchmark datasets, test scenarios, regression suites, and release quality gates.
Analyze agent behavior, failed trajectories, user feedback, and production telemetry to identify improvement opportunities.
Iterate on instructions, prompts, model selection, tool design, routing, context strategies, and orchestration.
Run controlled experiments and document measurable quality improvements.
Use Gemini models for reasoning, structured generation, code assistance, summarization, tool calling, and multimodal use cases.
Apply prompt engineering, structured outputs, grounding, safety controls, token and context optimization, and model-selection strategies.
Balance solution quality, latency, reliability, and cost across model and architecture choices.
Deploy and operate agents on GCP using appropriate services such as Vertex AI, Agent Runtime, Cloud Run, Google Kubernetes Engine, Cloud Functions, BigQuery, Pub/Sub, Cloud Storage, Cloud Monitoring, and Cloud Logging.
Implement CI/CD, infrastructure as code, automated testing, secrets management, IAM, and environment controls.
Apply AgentOps, LLMOps, MLOps, SRE, and cloud-native engineering practices to production AI systems.
Lead design reviews and establish reusable engineering standards and reference patterns.
Mentor engineers on ADK, Gemini, GCP, agent evaluation, AgentOps, and responsible AI practices.
Collaborate with product managers, architects, data scientists, security teams, and business stakeholders.
Communicate technical trade-offs, risks, dependencies, and outcomes to both technical and non-technical audiences.