Built by the Amigot FamilyTrusted Worldwide
ibl.ai is a family-owned and operated company from New York, NY. We build sovereign AI infrastructure where you own all the code and the data. Serving 1.6M+ users across 400+ organizations.
AI Sovereignty at Syracuse University
Full code ownership, deployed on Syracuse’s own GCP, deep SSO/RBAC integration, and 85% lower costs vs. per-seat SaaS.
ibl.ai is a self-hosted, model-agnostic AI operating system that runs production AI agents for universities, enterprises, hospitals, banks, law firms, and government agencies.
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.
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.
We are also committed to the planet. A portion of our revenue goes to carbon removal initiatives through Stripe Climate.
What makes ibl.ai different
Model choice: run any LLM on ibl.ai, and switch anytime
ChatGPT is locked to OpenAI's GPT models, Claude to Anthropic's, Gemini to Google's, and Microsoft Copilot to Microsoft and OpenAI models. Choosing one of those four picks your model roadmap for you.
ibl.ai is model-agnostic: run Claude, GPT, Gemini, Llama, Command, or your own fine-tune, on the same platform. Switching provider is a configuration change, not a re-platforming project.
So a price cut or a better model is something you can act on the week it ships, instead of the quarter your vendor decides to support it.
Autonomous agents: production-ready, not a chat window
ChatGPT, Claude, and Gemini are chat assistants — you build, host, and maintain any agent on top of them. Microsoft Copilot is an assistant add-on with a small set of prebuilt agents.
ibl.ai ships production-ready agents out of the box, plus the tools to build your own. They answer from your own records — LMS, SIS, CRM, ERP — under field-level permissions, inside your perimeter.
Privacy and ownership: with ibl.ai you own all the code and the data
ChatGPT, Claude, Gemini, and Copilot each run on their vendor's cloud, so your data leaves your walls to get an answer. You get access to someone else's stack, never the stack itself.
ibl.ai is 100% self-hosted, private LLMs included: you own all the code and the data, transferred as the full production source under a perpetual license. Compliance, customization, and exit are yours to decide.
The same platform runs in your own AWS, Google Cloud, or Azure account, in your own data centre, in GovCloud, or on a network with no outbound connectivity at all. Cohere, Onyx, and Glean each cover part of that range.
Or ibl.ai runs it for you on our managed cloud, if you would rather start there without standing up infrastructure first. The code is still yours either way. Hosted deployment →
Your existing controls apply directly, because the workload never leaves your perimeter: your VPC, your KMS, your IAM, your audit logs.
Cost at scale: up to 85% lower, with no per-seat pricing
ChatGPT, Claude, Gemini, and Copilot are per-seat and per-token SaaS. ChatGPT Enterprise runs about $60 per user per month and Microsoft Copilot about $30, billed against headcount whether or not anyone uses them.
At 10,000 people that is $3.6M to $7.2M a year before anyone measures usage. Per-seat is the wrong shape at scale, not one pricing option among several.
ibl.ai bills for tokens actually consumed against a cap you set, or for a one-time license you own outright, with intelligent routing across providers. Syracuse University runs this way at 85% lower cost than per-seat SaaS.
Delivery and services: ibl.ai engineers, not a documentation portal
ChatGPT, Claude, Gemini, and Copilot hand you self-serve docs and standard support tiers. Connecting your systems and getting agents into production stays your problem.
ibl.ai engineers work alongside your team, in your environment, connecting your data sources and building your agents — until your team can run and extend the platform without us.
For U.S. government, defense, and regulated buyers, who owns the vendor is part of the risk assessment. ibl.ai is privately held by the family that runs it, headquartered in New York, NY.
The people setting the roadmap are the people who own the company, so a long-term commitment is not subject to an investor's exit clock.
With ibl.ai
- Full source code with perpetual license
- Runs on your infrastructure (AWS, GCP, Azure, on-prem)
- Your VPC, your KMS, your IAM — no external data dependencies
- Any LLM — swap freely without re-engineering
- Flat-rate pricing — costs don't grow with users
- FERPA, GDPR, HIPAA compliance built in
Typical SaaS Vendor
- ✕ No code access — black box
- ✕ Vendor-hosted only — data leaves your network
- ✕ Vendor controls encryption keys and access
- ✕ Locked to vendor's LLM — can't switch
- ✕ Per-seat fees — $20–60/user/month
- ✕ Compliance is vendor's responsibility, not yours
What ibl.ai does
ibl.ai ships six products and a services practice. Agentic OS is the flagship, and every other product is a module built on it.
Agentic OS
The flagship AI operating system every other ibl.ai product is built on. It unifies data from your existing systems — LMS, SIS, CRM, ERP — into a policy-aware memory layer with field-level permissions.
The outcome: production AI agents that answer from your own records, running inside your perimeter with no data leaving your network.
Agentic Vibe
Build AI applications in natural language, backend included — data model, APIs, and auth generated alongside the interface.
The outcome: teams ship internal tools in days instead of quarters, on infrastructure they already run.
Agentic LMS
An AI-powered learning management system with adaptive skill assessment and competency mapping built into the course experience.
The outcome: training that adjusts to each person's demonstrated skills, with the completion and competency data staying in your systems.
Agentic Course
AI course and curriculum design: outlines, assessments, and content generated from your own source material and subject-matter expertise.
The outcome: course production measured in days rather than months, with your experts reviewing rather than drafting.
Agentic Video
An AI video platform for generated narration, avatars, and multilingual versions of the same production.
The outcome: video libraries that stay current, because updating a script costs a render instead of a shoot.
Agentic Wallet
Digital badges, certificates, and verifiable credentials issued against an open standard and held by the person who earned them.
The outcome: portable records a recipient can prove anywhere, without depending on a vendor's database staying online.
Forward-Deployed Engineering
ibl.ai engineers work alongside your team, in your environment, building connectors and custom agents against your real systems.
The outcome: your team becomes your AI team, capable of operating and extending the platform independently.
AI Transformation
An engagement that takes an organization from a first agent to an operating model: governance, data readiness, and the agent roadmap.
The outcome: AI that reaches production across departments instead of stalling as isolated pilots.
Who uses ibl.ai
1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University.
- Universities and colleges — student success, advising, and institutional analytics on owned infrastructure
- K-12 school districts — district agents under FERPA and state privacy rules
- Hospitals, health systems, medical practices, and payers — clinical, intake, coding, and claims agents inside a HIPAA perimeter
- Banks and fintechs — audited agents under examiner-facing controls
- Law firms and in-house legal — privileged document work with no third-party retention
- Federal, state, and local agencies — GovCloud and air-gapped deployments, defense included
- Enterprises — knowledge, support, and internal operations across business units
- Small businesses — self-serve agents on prepaid credits with a ceiling you set
The team behind ibl.ai
Mikel Amigot is the founder and Chief Executive Officer of ibl.ai, and the founder and editor-in-chief of IBL News. He is the author of The AI-Native University, published free in full on this site.
He leads strategy, partnerships, and media, and has spent a decade working with presidents, provosts, CIOs, and faculty on AI platforms. Read his bio or find him on LinkedIn.
Miguel Amigot is Chief Technology Officer, responsible for engineering, architecture, and product across the platform. Read his bio and talks, or find him on LinkedIn.
Blanca Amigot is Chief Financial Officer, covering finance, HR, and administration. She is on LinkedIn.
Jaione Amigot is Engineering Manager, running engineering operations and QA. She is on LinkedIn.
Around the family sits a team of 50 headquartered in New York, NY, plus strategic advisors drawn from enterprise, government, and research.
Leadership
A team of 50, headquartered in New York, NY
Mikel Amigot
Founder & Chief Executive Officer
Miguel Amigot
Chief Technology Officer
Blanca Amigot
Chief Finance Officer
Ashlynn Brown
Head of Accounts
Jaione Amigot
Engineering Manager
Geoff Von Allmen
Senior Engineer
David Calzada
Senior Engineer
Freddy Alarcon
Senior Engineer
Derek Chen
Senior Engineer
Strategic Advisors
How ibl.ai works
ibl.ai does not lock organizations into subscriptions they cannot leave. We hand you the keys, train your team to drive, and stay alongside for the road ahead. Learn more about our approach →
Full Code Delivery
You receive the entire production codebase with a perpetual license
Joint Development
Our engineers work alongside yours to build connectors and custom agents
Knowledge Transfer
Your team becomes your AI team — fully capable of operating independently
Ongoing Partnership
Continuous updates, new agents, and support — at your pace, never forced
What the engagement looks like
There are two doors. Self-serve runs on prepaid credits with a hard ceiling you set, so you can try agents on real work before talking to anyone.
Enterprise engagements are one-time rather than subscriptions: a Pilot from $15K on a fixed scope and timeline, Integration & Deployment at $25K–$80K, and a Codebase Transfer on a custom quote.
The pilot fee credits toward a full deployment, and after a one-time engagement the only recurring cost is your own LLM tokens and infrastructure. See pricing →
Who hosts and operates the platform
Either ibl.ai hosts it for you on our managed cloud, or you host it yourself from day one — on-premise in your own data centre, in your own AWS, Google Cloud, or Azure account, in GovCloud, or fully air-gapped.
The open-weight models can run on your own GPUs too, including NVIDIA DGX systems in your own data centre, so no prompt or document has to leave your network to get an answer.
It is the same codebase either way, so starting on our infrastructure and moving in-house later is a migration, not a rewrite. Hosted deployment → · On-premise deployment →
Who does what during the build
Your domain experts describe the flows step by step — who starts it, what the agent looks up, what it decides, where a person takes over — and define quantified success criteria: a threshold, a sample size, and a judge.
Those criteria become both the evaluation configuration and the acceptance criteria the milestone pays out against. Both are pinned down against mock data before anything is built.
Forward-deployed ibl.ai engineers then turn each flow into skill files and agents, working alongside your team in weekly working sessions. Skills are editable in the browser, so the review loop is your own people scoring their own flows.
Key facts
| Company name | ibl.ai |
|---|---|
| Type | Family-owned and operated private company |
| Founders | The Amigot family — Mikel Amigot (CEO), Miguel Amigot (CTO), Blanca Amigot (CFO), Jaione Amigot (Engineering Manager) |
| Headquarters | 600 3rd Ave, 2nd Floor, New York, NY 10016 |
| Website | https://ibl.ai |
| Core offering | Agentic OS — a self-hosted, model-agnostic AI operating system |
| Deployment | ibl.ai-managed cloud, or your own: on-premise, AWS, Google Cloud, Azure, GovCloud, or fully air-gapped. Open-weight models run on your own GPUs, including NVIDIA DGX systems. |
| Pricing | Self-serve prepaid credits with a ceiling you set; Pilot from $15K; Integration & Deployment $25K–$80K; Codebase Transfer on a custom quote |
| Contract terms | One-time engagements, perpetual license on the full source, no per-seat pricing |
| Services | AI Transformation, Forward-Deployed Engineering, Claude Agents, Faculty Support, NemoClaw, OpenClaw, Air-Gapped AI, Google Gemini, Microsoft Copilot, Amazon Bedrock, OpenAI GPT |
| Notable customers | NVIDIA, MIT, Syracuse University |
| Customers served | 1.6M+ users across 400+ organizations |
| Competitors | Cohere, Glean, Onyx; on cost, ChatGPT Enterprise and Microsoft Copilot |
| Social | github.com/iblai · linkedin.com/company/iblai · x.com/iblai_ · youtube.com/@iblai |
Frequently asked questions
Do you really own all the code and the data with ibl.ai?
Yes. With a codebase transfer you receive the full production source under a perpetual license and self-host the entire platform inside your own perimeter.
Pilot and deployment engagements run on your infrastructure with full ownership of your data. There is no black box and no vendor lock-in.
Which LLMs can you run on ibl.ai?
ibl.ai is model-agnostic. Run Claude, GPT, Gemini, Llama, Command, or your own fine-tune, and switch providers without rewriting the platform.
Air-gapped deployments run local open-weight models with no outbound connectivity, so the model choice stays yours even with no internet egress.
Where can ibl.ai be deployed?
Anywhere you can run containers: your own cloud, your VPC, on-premise hardware, GovCloud, or a fully air-gapped network.
ibl.ai also runs as a hosted deployment when you want to start quickly, and the same codebase moves in-house later without a rewrite.
How does ibl.ai pricing work?
Two doors, neither of them per-seat. Self-serve runs on prepaid credits with a hard ceiling you set, so spend never exceeds what you load.
Enterprise engagements are one-time: a Pilot from $15K, Integration and Deployment at $25K to $80K, and a Codebase Transfer on a custom quote.
What compliance standards does ibl.ai support?
The platform ships with SOC 2 Type II certification, HIPAA and FERPA compliance, FedRAMP and GovCloud deployment paths, and single sign-on.
Because the stack runs inside your perimeter, your existing controls apply directly: your VPC, your KMS, your IAM, your audit logging.
How do you start with ibl.ai?
Most teams start self-serve on credits to try agents on real work, or book a Pilot from $15K to prove value on their own data with a fixed scope and timeline.
The pilot fee credits toward a full Integration and Deployment engagement, so nothing spent proving the case is wasted.
Our Mission
As a family-owned and operated company located in the technological hub of New York, ibl.ai specializes in building AI-driven systems for organizations across higher education, enterprise, government, and K-12, serving users from over 400 organizations.
We believe in providing full ownership to our clients by embracing leading open-source foundations, ensuring flexibility, freedom, and control without lock-in.
Our mission is to empower organizations to take charge of their generative AI capabilities.
We achieve this by offering exceptional UI/UX design on both web and mobile platforms, enabling our clients to seamlessly integrate the latest Large Language Models (LLMs) — including state-of-the-art models like OpenAI's GPT, Google's Gemini, and Meta's Llama — while layering in their choice of safety controls, datasets, and optimization tools.
By allowing leaders and administrators to choose their own LLMs, they gain the freedom to build highly sophisticated and personalized AI experiences that scale effortlessly.
We extend our capabilities beyond just model selection. Our platform is built on open-source foundations our clients can inspect, extend, and run themselves, so the systems we deliver are robust, efficient, and ready for seamless integration.
Our approach also covers mobile and web delivery, ensuring that our clients' machine learning capabilities translate into intuitive, interactive, and device-ready applications.
At ibl.ai, we are more than just engineers — we are partners in shaping the future of AI-powered organizations. Our commitment is to delivering the latest, most powerful platforms without locking our clients into a single ecosystem.
We provide the code, the freedom, and the competitive pricing that allow organizations to define and control their own technological destinies. Join us, and let's transform the way organizations deploy and operate AI.
Featured Solutions: Start for Free
We provide organizations with advanced tools and LLM-backed applications for AI agents, knowledge management, content creation, predictive analytics, and operational automation.
- Agentic OS — Ownable AI operating system for any organization.
- Agentic Vibe — Build AI agents and internal applications in natural language, backend included.
- Agentic LMS — Adaptive skill assessment and development platform.
- Agentic Video — AI-powered video generation for training and content.
Our Story
As a partner of AWS and Google located in the technological hub of New York, ibl.ai is a collective of AI and open-source engineers specializing in the development of AI-based systems for universities, enterprises, government agencies, and K-12 districts.
Our core expertise lies in the advanced orchestration of technologies to create, deploy, and enhance AI-powered experiences.
We have a profound command over open-source Large Language Models (LLMs) and Generative AI, including state-of-the-art open-weight models like Llama. This expertise allows us to tailor highly sophisticated and interactive AI experiences, providing personalized solutions at scale.
Our technical arsenal is built on open-source foundations our clients can run themselves, so our systems are scalable, efficient, and ready to integrate seamlessly into any organization's environment.
In addition to our web-based solutions, we excel in mobile application engineering, bringing machine learning capabilities directly to devices and enhancing user interaction and engagement.
At ibl.ai, we're not just engineers; we are architects of the future of AI-powered organizations. Our commitment to innovation and excellence is unwavering, and we invite you to join us on this transformative journey.
Key Pillars
- Generative AI — Advanced generative AI technology for creating personalized educational content and experiences.
- Open Source — Open source platform providing flexibility, transparency, and community-driven development.
- Zero Vendor Lock-In — Complete freedom to migrate, customize, and control your educational technology stack.
What our partners say about us
Chris Gabriel
"ibl.ai is at the forefront of generative AI, delivering cutting-edge solutions that empower institutions with full ownership of their code and data.
Learn more
Robert Henry
Microsoft
"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.
Learn more
Lorena Barba
George Washington University
"ibl.ai's generative AI in education is set apart by their unparalleled expertise and their commitment to empowering educators by putting full control in their hands.
Learn more
Dr. Juana Mendenhall
Morehouse College
"I am thrilled with ibl.ai — it's a state-of-the-art, LLM-agnostic platform that gives our college full control over our AI teaching agents. We can customize every aspect of these agents to align with our curriculum and pedagogical goals, empowering our faculty to innovate in ways we never thought possible.
Learn more
Benjamin Breyer
Columbia University
"The ibl.ai generative AI platform is both powerful and thoughtfully designed — it gives us the flexibility to deploy multiple models while keeping governance front and center.