AI Architect - Generative AI Agents & Automation

About the role

10Pearls, a global digital engineering partner, is looking for a hands-on AI Architect to design and deliver AI agents and automation across the Microsoft ecosystem. The role covers both pro-code and low-code options, including Azure AI Foundry, Copilot Studio, Microsoft 365 Copilot and Power Platform, and expects the architect to choose the simplest option that meets the needs. The architect works directly with client stakeholders, leads technical discovery and contributes to the build within a small team.

Key responsibilities

• Solution architecture: define end-to-end architectures for AI agents, document processing and workflow automation.
• Platform selection: recommend configure, build or buy options across Foundry, Copilot Studio, Microsoft 365 Copilot, Power Platform and third-party tools.
• Cost modelling: forecast license, capacity and token costs for each use case, including long documents.
• Agent design: design single and multi-agent solutions with roles, routing, tool use and human approval steps.
• Document AI: design extraction, parsing and analysis for long documents such as contracts, including OCR and chunking.
• Knowledge grounding: work with the BA and SMEs to turn expert knowledge, policies and SOPs into structured content for agents.
• Microsoft 365 integration: integrate with Outlook, shared mailboxes, SharePoint, Teams and Microsoft Graph; assess access to line-of-business systems via APIs, data gateways or scheduled exports.
• AI security: design for prompt injection defense, file and content safety, access control, audit logging and data residency.
• Evaluation and quality: set up evaluation datasets, regression suites, confidence thresholds and accuracy tracking.
• Readiness and environments: confirm licenses, permissions, access and environment ownership across development stages.
• Hands-on delivery: contribute to prompts, agent configuration and code alongside AI and integration engineers.
• Pipeline preparation and stakeholder communication: assess upcoming use cases and explain trade-offs, cost and risk to stakeholders.

Required profile

• 8+ years in software engineering or solution architecture, with at least 2 years delivering generative AI solutions in production.
• Client-facing consulting or services delivery experience.
• Able to work regular overlap hours with US Eastern time.

Required skills

• Hands-on experience with Azure AI Foundry, including agents, model selection and evaluation.
• Copilot Studio and Power Automate.
• Good understanding of Microsoft 365 (Exchange, SharePoint, Teams, Graph, Entra ID).
• Working knowledge of Microsoft AI licensing (Copilot, Copilot Studio capacity, Azure consumption).
• Document extraction and analysis, including Azure Document Intelligence or similar.
• RAG, prompt design, agent orchestration and LLM evaluation.
• Understanding of AI security risks such as prompt injection and data leakage.
• Python sufficient to prototype and review code.