← All open roles
AI Growth Systems

AI Agent Developer – Copilot Studio & SharePoint

Remote (U.S.)
Part-time contract (1099)

9 North Group is hiring a remote, part-time AI Agent Developer to build Copilot Studio agents, SharePoint knowledge solutions, and Power Automate flows for agriculture and food companies, working across Claude, OpenAI, and Gemini. This is a 1099 independent contractor role, typically 10–20 hours per week depending on project load, with clearly scoped builds and no client-facing sales.

About 9 North Group

9 North Group is a commercialization and growth advisory firm helping innovative companies across the agriculture and food industries accelerate their path to market.

Through our expanding AI practice, we help companies implement practical AI solutions that improve productivity, streamline workflows, and strengthen commercial capabilities.

The opportunity

You'll be the hands-on technical delivery partner to the lead of our AI practice.

This is a pure engineering role. Our AI lead works with the client, maps the workflows, designs the solution, and defines the build scope. You take that scope and turn it into something that works: you build it, test it, document it, and hand it back ready for the client to run. We still want your technical judgment: if there's a simpler architecture, a better platform fit, or a risk in the design, say so before you build.

Most of the work is practical and fast-moving: light agents, low-code solutions, and SharePoint "intelligence layers" that help companies organize their knowledge, generate documents, capture expert know-how, and get trusted answers from their own content. Work is engaged project by project, with defined deliverables for each build, and you decide how and when it gets done.

Compensation is $85–$125/hr, depending on experience, with flexible hours and a paid work sample.

What you'll do

  • Build light agents: Build and test Copilot Studio and Microsoft 365 agents grounded in SharePoint content, including regulatory and technical document Q&A with clear citations.
  • Set up SharePoint intelligence layers: Implement the information architecture that makes client knowledge AI-ready: sites, libraries, content types, managed metadata, versioning, and approval workflows.
  • Deliver low-code solutions: Build Power Automate and agent flows for document generation, approvals, data collection, and QA/QC processes.
  • Build across AI platforms: Configure agents on Copilot Studio, Claude, OpenAI, and Gemini as each design calls for, and connect them to client data and tools through the Model Context Protocol (MCP).
  • Build agent harnesses: Implement the context, tools, rules, permissions, and evaluation loops that make agents reliable, repeatable, and safe to use in real workflows.
  • Implement knowledge graphs: Build entity and relationship models, such as products, regulations, crops, sites, and suppliers, and graph-based retrieval where the design calls for it.
  • Test and harden: Run question-set evaluations, fix what fails, and configure permissions, sensitivity labels, and environments to specification.
  • Document and hand off: Write clear build notes and runbooks so each solution can be handed to the client and sustained without you.
  • Give technical input on design: Review each build scope before you start, flag risks and limits, and suggest better technical approaches where you see them.
  • Build reusable assets: Turn repeated work into templates, schemas, and harness components that speed up the next build.

What we're looking for

  • Hands-on experience building Copilot Studio agents, ideally including at least one used by real people.
  • Strong SharePoint and Power Automate experience, including information architecture, content types, metadata, permissions, approvals, and document generation.
  • Practical, hands-on use of at least one non-Microsoft AI platform (Claude, OpenAI, or Gemini), with the willingness to build across all three.
  • A working understanding of how agents use knowledge, tools, and instructions, and how to test whether they give accurate, well-sourced answers.
  • Exposure to knowledge graphs, ontologies, or entity-relationship data modeling through projects, coursework, or hands-on experimentation.
  • The ability to deliver a written build scope accurately, on time, and with little supervision: testing your own work before you call it done, and communicating clearly through build notes, runbooks, and early warnings when a scope won't work as designed.
  • U.S.-based, with some availability for check-ins during Central Time business hours.

Bonus experience

  • Experience with Copilot in SharePoint, autofill columns, or structured document generation.
  • Experience with agent frameworks and SDKs (Claude Agent SDK, OpenAI Agents SDK, Google ADK, or Microsoft Agent Framework), MCP servers, or knowledge graphs (GraphRAG, Neo4j, Microsoft Fabric ontologies, or RDF/OWL).
  • Experience with website builders such as Framer, Webflow, or WordPress, or AI app builders such as Lovable.
  • Experience working with regulated, technical, or compliance-related documents.
  • Background in agriculture, food, life sciences, or related industries.
  • Previous consulting, Microsoft partner, fractional, or independent contractor experience.

What makes this opportunity different

  • Diverse work: Build across multiple clients, use cases, and AI platforms rather than maintaining a single system.
  • Grow your skills: Build depth in Microsoft 365 while expanding into Claude, OpenAI, Gemini, agent harnesses, and knowledge graphs on real client work.
  • A real technical voice: Your input shapes how solutions are designed and which reusable assets we build next.
  • Industry that matters: Your builds go to work inside real agriculture and food businesses, from emerging innovators to established industry leaders.
We're not looking for someone to sell AI or talk about it. We're looking for someone who wants to build it.

How to apply

Send a brief note describing one agent, SharePoint solution, or website you built: which platform you used, what challenges you ran into, and how you measured its success. Shortlisted candidates complete a short, paid work sample.

Interested? Email hello@9northgroup.com with the subject line “AI Agent Developer”.

Apply by email
{"@context": "https://schema.org", "@type": "JobPosting", "title": "AI Agent Developer – Copilot Studio & SharePoint", "description": "9 North Group is hiring a remote, part-time AI Agent Developer to build Copilot Studio agents, SharePoint knowledge solutions, and Power Automate flows for agriculture and food companies, working across Claude, OpenAI, and Gemini. This is a 1099 independent contractor role, typically 10–20 hours per week depending on project load, with clearly scoped builds and no client-facing sales.\nAbout 9 North Group\n9 North Group is a commercialization and growth advisory firm helping innovative companies across the agriculture and food industries accelerate their path to market.\nThrough our expanding AI practice, we help companies implement practical AI solutions that improve productivity, streamline workflows, and strengthen commercial capabilities.\nThe opportunity\nYou'll be the hands-on technical delivery partner to the lead of our AI practice.\nThis is a pure engineering role. Our AI lead works with the client, maps the workflows, designs the solution, and defines the build scope. You take that scope and turn it into something that works: you build it, test it, document it, and hand it back ready for the client to run. We still want your technical judgment: if there's a simpler architecture, a better platform fit, or a risk in the design, say so before you build.\nMost of the work is practical and fast-moving: light agents, low-code solutions, and SharePoint \"intelligence layers\" that help companies organize their knowledge, generate documents, capture expert know-how, and get trusted answers from their own content. Work is engaged project by project, with defined deliverables for each build, and you decide how and when it gets done.\nCompensation is $85–$125/hr, depending on experience, with flexible hours and a paid work sample.\nWhat you'll do\n• Build light agents: Build and test Copilot Studio and Microsoft 365 agents grounded in SharePoint content, including regulatory and technical document Q&A with clear citations.\n• Set up SharePoint intelligence layers: Implement the information architecture that makes client knowledge AI-ready: sites, libraries, content types, managed metadata, versioning, and approval workflows.\n• Deliver low-code solutions: Build Power Automate and agent flows for document generation, approvals, data collection, and QA/QC processes.\n• Build across AI platforms: Configure agents on Copilot Studio, Claude, OpenAI, and Gemini as each design calls for, and connect them to client data and tools through the Model Context Protocol (MCP).\n• Build agent harnesses: Implement the context, tools, rules, permissions, and evaluation loops that make agents reliable, repeatable, and safe to use in real workflows.\n• Implement knowledge graphs: Build entity and relationship models, such as products, regulations, crops, sites, and suppliers, and graph-based retrieval where the design calls for it.\n• Test and harden: Run question-set evaluations, fix what fails, and configure permissions, sensitivity labels, and environments to specification.\n• Document and hand off: Write clear build notes and runbooks so each solution can be handed to the client and sustained without you.\n• Give technical input on design: Review each build scope before you start, flag risks and limits, and suggest better technical approaches where you see them.\n• Build reusable assets: Turn repeated work into templates, schemas, and harness components that speed up the next build.\nWhat we're looking for\n• Hands-on experience building Copilot Studio agents, ideally including at least one used by real people.\n• Strong SharePoint and Power Automate experience, including information architecture, content types, metadata, permissions, approvals, and document generation.\n• Practical, hands-on use of at least one non-Microsoft AI platform (Claude, OpenAI, or Gemini), with the willingness to build across all three.\n• A working understanding of how agents use knowledge, tools, and instructions, and how to test whether they give accurate, well-sourced answers.\n• Exposure to knowledge graphs, ontologies, or entity-relationship data modeling through projects, coursework, or hands-on experimentation.\n• The ability to deliver a written build scope accurately, on time, and with little supervision: testing your own work before you call it done, and communicating clearly through build notes, runbooks, and early warnings when a scope won't work as designed.\n• U.S.-based, with some availability for check-ins during Central Time business hours.\nBonus experience\n• Experience with Copilot in SharePoint, autofill columns, or structured document generation.\n• Experience with agent frameworks and SDKs (Claude Agent SDK, OpenAI Agents SDK, Google ADK, or Microsoft Agent Framework), MCP servers, or knowledge graphs (GraphRAG, Neo4j, Microsoft Fabric ontologies, or RDF/OWL).\n• Experience with website builders such as Framer, Webflow, or WordPress, or AI app builders such as Lovable.\n• Experience working with regulated, technical, or compliance-related documents.\n• Background in agriculture, food, life sciences, or related industries.\n• Previous consulting, Microsoft partner, fractional, or independent contractor experience.\nWhat makes this opportunity different\n• Diverse work: Build across multiple clients, use cases, and AI platforms rather than maintaining a single system.\n• Grow your skills: Build depth in Microsoft 365 while expanding into Claude, OpenAI, Gemini, agent harnesses, and knowledge graphs on real client work.\n• A real technical voice: Your input shapes how solutions are designed and which reusable assets we build next.\n• Industry that matters: Your builds go to work inside real agriculture and food businesses, from emerging innovators to established industry leaders.\nWe're not looking for someone to sell AI or talk about it. We're looking for someone who wants to build it.\nHow to apply\nSend a brief note describing one agent, SharePoint solution, or website you built: which platform you used, what challenges you ran into, and how you measured its success. Shortlisted candidates complete a short, paid work sample.", "datePosted": "2026-11-20", "employmentType": ["CONTRACTOR", "PART_TIME"], "hiringOrganization": {"@type": "Organization", "name": "9 North Group", "sameAs": "https://www.9northgroup.com", "url": "https://www.9northgroup.com"}, "jobLocationType": "TELECOMMUTE", "applicantLocationRequirements": {"@type": "Country", "name": "United States"}, "industry": "Agriculture and Food", "directApply": false, "url": "https://www.9northgroup.com/careers/ai-agent-developer", "baseSalary": {"@type": "MonetaryAmount", "currency": "USD", "value": {"@type": "QuantitativeValue", "minValue": 85, "maxValue": 125, "unitText": "HOUR"}}}