We are looking for a visionary Solution Architect to lead the design and delivery of next-generation Agentic AI systems on Google Cloud Platform (GCP). In this role, you will architect a sophisticated Deep Research Multi-Agent System, where specialized AI agents (Researchers, Critics, Synthesizers, Reviewers) autonomously collaborate to solve open-ended problems, analyze large datasets, and produce comprehensive insights.

This role goes beyond building simple chatbots - you will create System 2 style reasoning applications using Vertex AI Agent Builder, Gemini 1.5 Pro/Flash, and advanced orchestration frameworks such as LangGraph or Googles Agent Development Kit (ADK).

Key Responsibilities :

1. Multi-Agent Architecture Design :

- Design a GCP-native multi-agent architecture that supports deep-research workflows : planning, exploration, self-correction, and synthesis.

- Define agent roles and collaboration patterns (Hierarchical, Sequential, Joint-Collaboration) using the Agent-to-Agent (A2A) Protocol.

- Build robust state-management systems using Cloud SQL, AlloyDB, or Redis to maintain long-running agent state, memory banks, and thread history.

2. GCP-Native Development & Integration :

- Lead implementation using Vertex AI Agent Builder and Vertex AI Reasoning Engine.

- Integrate Gemini 1.5 Pro (1M+ token context window) to process large datasets efficiently with long-context caching.

- Build tool-based function calling pipelines connecting agents to BigQuery, Cloud Run APIs, and enterprise knowledge sources like Vertex AI Search.

3. Advanced Orchestration & Reasoning :

- Develop complex orchestration flows (loops, branching, conditionals) using LangGraph or Google ADK.

- Implement Human-in-the-Loop (HITL) checkpoints for high-stakes decisions.

- Create reflection & critique pipelines where reviewer agents validate outputs from worker agents for factual accuracy.

4. Production Readiness & Governance :

- Establish robust LLM Ops practices : evaluation with AutoSxS, tracing with OpenTelemetry, and structured prompt management.

- Implement strong security guardrails using Vertex AI Model Armor or custom filters.

- Optimize model usage by balancing Flash models (reasoning) and Pro models (synthesis).

Required Qualifications :

- Bachelors or Masters degree in Computer Science, AI, Data Science, or related fields.

- 7+ years of experience as a Solution Architect, with 3+ years focused on Google Cloud Platform (GCP).

- Strong hands-on experience with LLMs (Gemini, PaLM, GPT-4) and clear understanding of Agents vs. Chains.

- Experience with agent frameworks such as LangChain, LangGraph, Google ADK, or CrewAI.

Technical Skills :

i. Languages : Python (Expert), TypeScript (Preferred)

ii. GCP Services : Vertex AI, Cloud Run, Cloud Functions, BigQuery, Pub/Sub, Firestore/AlloyDB

iii. Vector Search : Vertex AI Vector Search, milvus

- Google Cloud certifications such as :

1. Professional Cloud Architect

2. Professional Machine Learning Engineer

Preferred Qualifications (Deep Research Specialist) :

- Hands-on experience building systems capable of multi-step reasoning- breaking vague prompts into structured tasks (search - extract - compare - synthesize - report).

- Understanding of standardized Agent-to-Agent (A2A) Protocols for cross-framework interoperability.

- Experience optimizing workflows for 1M+ token context windows, including handling needle-in-a-haystack challenges.