Maranatha Poirier

Hi, I'm Maranatha

From data center to AI, I make technology work.

I've spent twenty years building and running the systems organizations depend on. My current focus is AI, and those years are what let me help organizations adopt it without losing control of identity, data or governance.

  • Infrastructure
  • Cloud platforms
  • Administration
  • Security
  • Integration
  • AI current focus
Portrait of Maranatha Poirier, smiling, wearing glasses, a navy blazer and a light blue shirt
Now: enterprise AI enablement
20+years in IT
1,200+people moved to SSO and SCIM provisioning
300+servers refreshed across VMware, Azure and AWS
99.9%uptime for 18 clients over a decade

How I think about the work

Behind every login screen is someone who just wants to get to their work. I want them in on the first try, with access to what they need and nothing they shouldn't have, and never a thought about who set it up.

I've scoped and rolled out technology for organizations of every size, from a managed IT practice with 18 clients to a 300-server hybrid-cloud environment at a Texas state agency. Now I point that experience at AI enablement: agent deployments, connector governance, and the identity and data work that makes enterprise AI safe to use day to day. I still build things myself, including SCIM provisioning, Entra ID and Okta integrations, Microsoft Fabric data pipelines, and a growing set of AI products.

The work

What the work looks like

A job-by-job résumé would tell you less than the problems I get called in to fix and what changed afterward.

NowEnterprise AI enablement

Building AI into how organizations already work

I design agent deployments, connector governance and data pipelines for organizations adopting AI at scale. One recent build was a Microsoft Fabric data factory on a three-tier medallion architecture, with Spark notebooks and Dataflow Gen2 doing the heavy lifting. I'm growing this into an independent practice called AIVENY.

1,200people on SSO and SCIM

Copilot and identity rollout across a dozen departments

For a large public-sector organization I led the Microsoft 365 Copilot rollout, covering provisioning, connector configuration, data access and adoption. That meant integrating Entra ID, Okta and Intune, running SCIM provisioning and SSO for a workforce of more than 1,200, and refreshing a hybrid-cloud environment of 300+ servers across VMware, Azure and AWS.

18clients, 27 locations

An independent IT practice with no incidents to report

For over a decade I ran managed IT for 18 small and mid-sized businesses across 27 locations. I handled networking, servers, storage, security, cloud, websites, backups and staff training. Uptime held at 99.9%, with zero data loss and zero malware incidents, year after year.

60%faster device provisioning

Modernizing infrastructure and automating the tedious parts

I rebuilt data center and endpoint infrastructure for large return-to-office transitions and wrote the PowerShell behind zero-touch Windows deployment. Earlier in my career I took a growing company's IT footprint from one site to eight, which quadrupled transaction throughput and opened up several million dollars in new annual revenue.

75%fewer client incidents

Leading a technical team for nonprofits

I led a team supporting nonprofit organizations and cut client incidents by 75% and licensing costs by 80% through platform consolidation and technology grants. I've brought the same habits to my own ventures since.

Toolkit

What's in the toolkit

AI and agent platforms

  • Anthropic Claude Enterprise, Claude Code and Skills
  • Google Gemini
  • Microsoft 365 Copilot and Copilot Studio
  • OpenAI ChatGPT, Codex and Agents SDK
  • MCP
  • LangChain and LangGraph
  • CrewAI

AI engineering

  • Prompt engineering and optimization
  • Agent workflow design and deployment
  • Microsoft Fabric
  • RAG pipeline design
  • llama.cpp and vLLM hosting

Cloud and platforms

  • Azure
  • AWS
  • Google Cloud
  • VMware vSphere, ESXi and vCenter 8
  • Docker
  • Hyper-V
  • Windows Server and Linux
  • SharePoint and Teams

Identity, SaaS and automation

  • Entra ID
  • Okta
  • OAuth 2.0
  • Active Directory and GPO
  • PowerShell and Python
  • Apache Spark
  • n8n and Zapier
  • GitHub and CI/CD

If you're adopting AI and want to keep control of identity and data, let's talk.

Tell me what you're working with and where it hurts. I'll tell you what I'd do first.