Senior GenAI Developer (Agentic AI Specialist)

Job Title: Senior GenAI Developer (Agentic AI Specialist)
Location: Remote

Duration: 12 Months

Reworked Job Summary

The Senior GenAI Developer will architect and deploy production-grade Agentic AI systems using Google Vertex AI Agent Builder and Google Agent Development Kit (ADK). This hands-on role focuses on autonomous agents that execute complex business workflows through tool calling, API orchestration, advanced RAG, and robust state management within the Google Cloud ecosystem.

Core Responsibilities

Agent Engineering & Orchestration

• Design sophisticated multi-agent systems using Vertex AI Agent Builder and Google ADK for end-to-end workflow automation.
• Deploy and manage agent lifecycles with Vertex AI Agent Engine, implementing scaling, versioning, and rollback strategies.
• Implement tool calling patterns integrating enterprise APIs, databases, and custom functions for autonomous task execution.

Advanced RAG Architecture

• Architect high-fidelity RAG pipelines using Google RAG Engine with hybrid retrieval (semantic + keyword) and intelligent reranking.
• Optimize embedding strategies and context chunking for financial/legal/technical domain accuracy.
• Implement multi-hop retrieval combining structured GraphRAG with unstructured document search.

State & Memory Management

• Design long-term context persistence using Google Memory Bank for multi-session user continuity and agent memory.
• Implement conversation state machines handling branching workflows, user interruptions, and context handoff.

Security & Governance

• Deploy Google Model Armor and Google Guardrails preventing prompt injection, PII leakage, and toxic content generation.
• Configure enterprise-grade access controls, audit logging, and compliance monitoring for production AI deployments.

Evaluation & Benchmarking

• Build Golden Datasets and implement LLM evaluation frameworks measuring agent task success, hallucination rates, and response fidelity.
• Establish continuous evaluation pipelines with A/B testing and production monitoring.

Technical Requirements

Google Cloud AI Stack

Vertex AI Suite:

• Agent Builder: Multi-agent orchestration
• Agent Engine: Production deployment/scaling
• RAG Engine: Enterprise retrieval pipelines
• Memory Bank: Long-term agent context
• Model Armor/Guardrails: Security hardening

Google ADK:

• Tool calling frameworks
• Workflow patterns (ReAct, CoT)
• Agent lifecycle management

Technical Proficiencies

• Languages: Python (expert), async/await patterns
• AI Frameworks: LangChain/LangGraph integration
• Evaluation: Ragas, DeepEval, custom agent benchmarks
• Infra: Kubernetes, GKE, Cloud Run for AI workloads
• Monitoring: Cloud Monitoring, Logging, Trace sampling

Key Competencies

Agentic AI Patterns

• ReAct (Reasoning + Acting) loops
• Chain-of-Thought orchestration
• Multi-agent collaboration patterns
• Tool calling + API chaining
• Human-in-the-loop escalation

Production Excellence

• Agent observability + debugging
• Error recovery + fallback strategies
• Cost optimization (token limits, caching)
• SLO monitoring (latency, success rate)
• A/B testing + champion/challenger

Keywords: Senior GenAI Developer, Agentic AI, Vertex AI Agent Builder, Google ADK, Agent Engine, Google RAG Engine, Memory Bank, Model Armor, Google Guardrails, Multi-Agent Systems, Tool Calling, ReAct Pattern, Chain of Thought, RAG Pipeline, LLM Evaluation, Golden Datasets, Python Async, Production AI, Autonomous Agents, API Orchestration, Workflow Automation, Enterprise AI Security, Cloud Monitoring, A/B Testing AI, Remote GenAI

About VDart Group
VDart Group is a global leader in technology, product, and talent solutions, serving Fortune 500 clients in 13 countries. With over 4,000 professionals worldwide, we deliver innovation, operational excellence, and measurable outcomes across industries. Guided by our commitment to People, Purpose, and Planet, VDart is recognized with an EcoVadis Bronze Medal and as a UN Global Compact member, reflecting our dedication to sustainable practices.