Qureos

Find The RightJob.

GenAI / Agentspace Engineer

GenAI / Agentspace Engineer — Federal Cloud Modernization

Remote (US) · Contract · $130–$155/hr DOE

A Google Cloud Partner is hiring a GenAI / Agentspace Engineer to lead the Phase 2 rollout of Gemini-powered AI search and conversational AI for a major IRS modernization program. You'll build the layer that agency users actually interact with — AI-powered document search, summarization, and question answering over millions of federal documents.

This is the visible tip of the spear. Everything the infrastructure, networking, and ML teams are building feeds into what you create. If you want to be the person who makes Gemini work at federal scale, this is the role.

What you'll actually do:
— Lead Agentspace pilot rollout and configuration
— Integrate Gemini API for conversational search and document Q&A
— Build AI-powered summaries with citation, grounding, and attribution
— Engineer prompts to minimize hallucination in a compliance environment
— Design agent orchestration patterns for multi-step retrieval
— Run user acceptance testing with agency stakeholders

What you need (non-negotiable):
— GCP Professional ML Engineer certification or Google GenAI specialty
— 3+ years GenAI / LLM engineering (Gemini, PaLM, GPT, or Claude)
— Enterprise AI search deployment (Agentspace, Vertex AI Search, or equivalent)
— Prompt engineering, grounding, and hallucination mitigation in production
— US Citizen
— IRS MBI clearance (active) or ability to obtain

What puts you ahead:
— Google Agentspace deployment experience (rare and valuable)
— Federal conversational AI implementations
— RAG + agent orchestration patterns (LangChain, LlamaIndex, or custom)
— Evaluation frameworks for GenAI (faithfulness, relevance, toxicity)

12-month contract · Fully remote (US only) · Lead the AI layer that users actually see and touch

Pay: $130.00 - $155.00 per hour

Application Question(s):

  • Do you hold a GCP ML Engineer cert or Google GenAI certification? [Yes/No]

Security clearance:

  • Confidential (Preferred)

Work Location: Remote

© 2026 Qureos. All rights reserved.