ibl.ai Agentic AI Blog

Insights on building and deploying agentic AI systems. Our blog covers AI agent architectures, LLM infrastructure, MCP servers, enterprise deployment strategies, and real-world implementation guides. Whether you are a developer building AI agents, a CTO evaluating agentic platforms, or a technical leader driving AI adoption, you will find practical guidance here.

Topics We Cover

Featured Research and Reports

We analyze key research from leading institutions and labs including Google DeepMind, Anthropic, OpenAI, Meta AI, McKinsey, and the World Economic Forum. Our content includes detailed analysis of reports on AI agents, foundation models, and enterprise AI strategy.

For Technical Leaders

CTOs, engineering leads, and AI architects turn to our blog for guidance on agent orchestration, model evaluation, infrastructure planning, and building production-ready AI systems. We provide frameworks for responsible AI deployment that balance capability with safety and reliability.

Back to Blog

ibl.ai on Microsoft Surface Copilot+ PCs: Local AI Tutoring Powered by the NPU

Miguel AmigotFebruary 13, 2026
Premium

ibl.ai runs directly on Microsoft Surface Copilot+ PCs, using the built-in Neural Processing Unit (NPU) to deliver real-time AI tutoring and content tools without requiring a cloud connection. Students get instant, on-device mentoring; faculty get powerful authoring tools; and institutions keep every byte of data local.

Not every campus has reliable broadband. Not every student wants their tutoring data traversing the public internet. And not every institution is comfortable routing sensitive academic interactions through a third-party cloud.

ibl.ai now runs natively on Microsoft Surface Copilot+ PCs, tapping into the device's built-in Neural Processing Unit (NPU) to deliver AI tutoring, content generation, and learning analytics—entirely on-device.


What Is a Copilot+ PC—and Why Does the NPU Matter?

Microsoft's Copilot+ PC initiative defines a new class of Windows devices equipped with dedicated AI accelerators. Surface Copilot+ PCs—including the Surface Pro and Surface Laptop lines—ship with NPUs capable of 40+ TOPS (trillion operations per second), powered by Qualcomm Snapdragon X Elite or Intel Lunar Lake processors.

The NPU is purpose-built for the matrix math that drives AI inference. Unlike the CPU (general-purpose) or GPU (graphics-first), the NPU runs AI models with dramatically lower power draw and heat output. That means:

  • Sustained performance. AI inference runs for hours on battery without thermal throttling.
  • No network dependency. The model lives on the device. No cloud round-trip, no latency spikes, no outage risk.
  • Data stays local. Student prompts, responses, and session logs never leave the machine unless the institution explicitly syncs them.

How ibl.ai Leverages the NPU

ibl.ai's platform is architecturally model-agnostic and deployment-flexible. The same AI agents—ibl.ai for tutoring, courseAI for content authoring, skillsAI for competency mapping—that run on cloud infrastructure can also run on-device using optimized small language models (SLMs) compiled for the Windows AI runtime.

Here is what happens when a student opens ibl.ai on a Surface Copilot+ PC:

  1. Model loads into NPU memory. A quantized SLM (e.g., Phi-3, Llama 3 8B, or a fine-tuned institutional model) is loaded onto the NPU at login. Load time is under two seconds.

  2. Tutoring session runs locally. The student asks a question. ibl.ai's orchestration layer routes the query to the on-device model. Response generation happens at 20–40 tokens per second—fast enough for a conversational experience that feels indistinguishable from cloud AI.

  3. RAG with local documents. Course materials, syllabi, and lecture notes stored on the device (or on a campus share) are indexed locally. The AI agent retrieves relevant passages and grounds its answers—no cloud vector database required.

  4. Sync when ready. Session summaries and learning analytics can optionally sync to the institution's cloud tenant (Azure, AWS, or ibl.ai's hosted platform) when the device reconnects—on the institution's terms and schedule.


A Vision for Institutional Control

Robert Henry, who leads Microsoft Surface for Education partnerships, described the collaboration:

"What set ibl.ai apart for us is their commitment to institutional control. Their platform taps into the NPU (neural processing unit) on Microsoft Surface Copilot+ PCs to offer offline AI experiences that feel as fast as the cloud. Students get instant, on‑device tutoring and faculty gain powerful content tools — all with data staying local. This collaboration reflects exactly the kind of innovation Microsoft Surface aims to accelerate in education."

That phrase—institutional control—is the operative concept. Universities, not vendors, decide:

  • Which models run on-device. Institutions can deploy Microsoft Phi-3 for general tutoring, a fine-tuned domain model for nursing or engineering, or an open-weight model from the ibl.ai model library.
  • What data stays local vs. syncs. Policies are configurable per department, per course, or per device fleet.
  • Who has access. Azure Entra ID (Azure AD) and institutional MDM policies govern device enrollment and AI feature access.

Use Cases That Come Alive On-Device

Fieldwork and Clinical Rotations

Nursing students on hospital rotations, education majors in K-12 classrooms, agriculture students in rural field stations—these learners are often offline or on restricted networks. With ibl.ai on Surface, they have an AI tutor in their bag that works without Wi-Fi.

Exam Prep in Locked-Down Mode

During proctored study sessions, institutions can restrict network access while still allowing on-device AI tutoring. The student interacts with ibl.ai; the device logs the session; no data leaves the machine until the proctor authorizes sync.

Faculty Content Authoring

Professors use courseAI on their Surface to draft quiz questions, generate lecture outlines, or create adaptive learning paths—all from their office, their couch, or an airplane. The NPU handles inference; the content exports to the LMS when ready.

Low-Bandwidth Campuses

Community colleges, satellite campuses, and institutions in developing regions often face connectivity constraints. On-device AI eliminates the cloud bottleneck and ensures every student has the same AI experience regardless of bandwidth.


Deployment: Simpler Than You Think

ibl.ai's Surface deployment follows the same MDM (Mobile Device Management) patterns IT teams already use for Windows devices:

  1. Package the ibl.ai app as an MSIX or Win32 package in Microsoft Intune.
  2. Push the AI model as a companion payload—or let the app download it once on campus Wi-Fi.
  3. Configure policies via Intune: model selection, sync schedule, data retention, feature flags.
  4. Monitor fleet health through the ibl.ai admin console, which aggregates anonymized usage metrics across the device fleet.

No GPU clusters. No VPN tunnels. No special hardware beyond the Surface itself.


Cloud + Edge: A Hybrid Strategy

On-device AI does not replace the cloud—it extends it. Institutions can run ibl.ai in the cloud for campus-wide analytics, heavy-duty research workflows, and models that require 70B+ parameters. The Surface NPU handles the everyday tutoring, content authoring, and offline scenarios.

ibl.ai's orchestration layer manages this hybrid routing transparently. When the device is online and the query exceeds the local model's capability, it escalates to the cloud. When offline, it stays local. The student never notices the difference.


The Bottom Line

Microsoft Surface Copilot+ PCs put a dedicated AI processor in every student's hands. ibl.ai puts a world-class AI tutor on that processor. Together, they deliver fast, private, always-available AI education—no cloud required, no data compromises, no infrastructure headaches.

Want to pilot ibl.ai on your Surface fleet? Contact us to get started.

Related: What Is an Enterprise LLM Platform? The One You Own ¡ Who Owns Your Data When You Use ChatGPT or Copilot?

Why does owning the AI stack matter?

ibl.ai is the agentic AI platform where you own all the code and the data. You self-host the entire stack inside your own perimeter, run it model-agnostic across any LLM and switch anytime, and pay by usage with no per-seat pricing — so you can deploy anywhere: your cloud, on-premise, GovCloud, or fully air-gapped.

  • You own all the code and the data

    Full source code under a perpetual license, running on your infrastructure. Not API access to someone else's platform — the stack itself is yours.

  • Model-agnostic

    Run any LLM — Claude, GPT, Gemini, Llama, Command, or your own fine-tune — and switch providers without rewriting the platform.

  • No per-seat pricing

    Usage-based billing against a budget cap you set. Cost tracks what your organization actually uses, not how many people you employ.

  • Deploy anywhere

    Your cloud, your VPC, on-premise, GovCloud, or a fully air-gapped network with no outbound connectivity.

1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University.

ibl.ai is family-owned and operated from New York, NY — a U.S.-headquartered, domestically-owned long-term partner, not a vendor that sells licenses and moves on.

See the ibl.ai AI Operating System in Action

Discover how leading universities and organizations are transforming education with the ibl.ai AI Operating System. Explore real-world implementations from Harvard, MIT, Stanford, and users from 400+ institutions worldwide.

View Case Studies
Work with our team

Pilots, deployment, and full ownership

Most enterprise engagements are one-time, not subscriptions. You integrate ibl.ai with your own data, deploy it on your own infrastructure, and the engineering hours scale with the work — so the price tracks the scope, not your headcount.

Start here

Pilot

from $15K

fixed scope ¡ fixed timeline

A time-boxed proof of value on your real data — not a slide deck.

Best for: Teams that want to see ibl.ai working before committing.

  • Deployed on your infrastructure or our cloud
  • 1–2 production agents wired to a slice of your data
  • One integration (LMS / SIS / SSO / data source)
  • Weekly working sessions with our engineers
  • Pilot fee credits toward a full engagement
Scope a pilot
Most common

Integration & Deployment

$25K – $80K

one-time ¡ not a subscription

Full deployment integrated with your data and systems. Engineering hours scale with scope.

Best for: Organizations rolling ibl.ai out across a department, campus, or business unit.

  • Platform deployed in your VPC, on-prem, or air-gapped
  • Integrated with your data + identity (SSO / SAML)
  • Multiple custom agents built to your workflows
  • Engineering hours proportional to scope
  • You own the data ¡ run any LLM you choose
Plan a deployment
Full ownership

Codebase Transfer + Custom AI Engineering

Six figures

perpetual license ¡ you own the stack

We transfer the full source code. You own and self-host the entire platform — outright.

Best for: Government, defense, and enterprises that require perpetual ownership and sovereignty.

  • Complete source-code transfer + perpetual license
  • Dedicated AI engineering team on your roadmap
  • Custom agents, models, and integrations to spec
  • Air-gapped capable ¡ zero vendor lock-in
  • Family-owned, New York–based long-term partner
Talk about ownership
You own the code and data Run any LLM — Claude, GPT, Gemini, Llama Family-owned & operated from New York, NY