We are seeking a Senior AI Software Engineer to build and scale an autonomous "Multi-Framework Strategy Engine" designed for enterprise advisory and consulting. You will not be building simple chatbots or wrappers. Instead, you will be designing complex, multi-agent orchestrations using the Google Ecosystem (Gemini / Agent Development Kit (ADK)) and Microsoft Ecosystem (Copilot).
Your primary objective is to support our Global Head of AI and Data for our Business unit to design, build and scale our agent architecture into a secure, stateless, multi-BU / multi-tenant platform capable of integrating with enterprise ecosystems (like Google Workspace and Microsoft Copilot) and securely querying private corporate data (RAG), as well as other internal corporate assets.
What You Will Do:
- Agentic Orchestration: Design, prompt, and tune specialised autonomous agents (Orchestrators, Researchers, Evaluators) using the Google ADK. You will optimise their conversational memory, tool-calling reliability, and "Introspection/Self-Correction" loops.
- Distributed SaaS Architecture: Design and implement a distributed enterprise-grade agentic workflow system within the Google environment.
- Enterprise RAG & Data Security: Implement/integrate to a secure "Datastore Router" pattern and other Vertex capabilities such as Enterprise Search / "Deep Research, strictly enforced by corporate security and data specifications.
- Stateful UI/UX Integration: Build / Integrate with event-driven React frontend, Copilot, Gemini interfaces.
- Ecosystem Integration: Architect our API layer so our specialised agentic workflows can be exposed as native "Skills" or "Plugins" within enterprise chat interfaces like Google Gemini Advanced and Microsoft Copilot.
Required Technical Skills
- Communication: Fluent English speaker with excellent technical writing and verbal communication skills.
- Expert Python (Backend): Deep experience with asynchronous Python (asyncio), FastAPI, and building high-throughput, decoupled systems (Celery, Redis, RabbitMQ).
- LLM & Agent Frameworks: Hands-on experience building *autonomous* workflows (not just RAG). Familiarity with Google ADK, LangChain, LlamaIndex, or AutoGen. Crucially, an understanding of context window management, token optimisation, and complex system prompting is required.
- Experienced in Database & Multi-Tenancy design and build: Strong SQL skills (PostgreSQL) and experience with SQLAlchemy and Alembic. Experience designing secure SaaS architectures (Row-Level Security, OAuth, API Key management).
- Cloud Infrastructure (GCP preferred): Experience deploying stateless applications to Kubernetes (GKE) or Cloud Run. Familiarity with Cloud Storage (GCS) and Google Vertex AI endpoints.
Nice-to-have technical Skills
- Front-end: Experience with React, Vite, and managing complex streaming states in browser, TSX, NotebookLM liked FE design/implementation/integration experience.
- RAG Expertise: Experience building and tuning vector databases and extraction pipelines (chunking, embeddings, OCR).
Non-technical skills
- Familiarity with consulting frameworks (Porter’s 5 Forces, SWOT etc) to better align agent prompts with strategic outcomes.
- Ability to communicate and present technical language, into simple language to engage the rest of the business
- Ability to work well across layers of the organisation (with Ipsos’ technical experts in AI and IT, the business unit central team, include operations support and the global business unit lead)
- Ability to show up well and engage in conversations as part of a global AI board, which includes the business unit central team and end-users of the agents you will be building.