Colleague AI Becomes First K-12 Platform to Empower Districts to Build Their Own Governed AI Tools and Assistants

New features let districts, schools and teachers build and customize AI tools around their policies, curriculum and instructional needs.

SEATTLE, WA, UNITED STATES, October 23, 2026 /EINPresswire.com/ — Colleague AI, the research-backed K-12 AI platform backed by $19M+ in federal research grants from the NSF, DOE, and IES, today became the first K-12 platform to let districts, schools and teachers build, clone and customize the AI Assistants and tools that power it, rather than choosing among a fixed set of pre-built tools.

Engineering, marketing and support teams have spent years building their own AI tools on top of enterprise platforms. K-12 has not had the same option — no platform combined the governance, the pedagogy and the education-specific architecture required to do it responsibly. Colleague AI’s “one platform, every stakeholder” model long meant a set of pre-built AI Assistants for teachers, students, schools and districts; the framework behind those tools is now visible and editable at the district, school and teacher level. A district can see the AI Assistants behind its own platform, clone one or build a new one, and shape it around a policy, a curriculum or a student population that no pre-built template quite fits.

“This is the moment K-12 catches up to where other industries already are — without asking a single teacher to become an engineer to get there,” said Dr. Min Sun, co-founder and CEO of Colleague AI. “For the first time, a district, a school and a teacher can build the AI tools they actually need, inside a platform built for education from day one, without ever leaving the research and governance foundation their board already trusts.”

“This is agentic architecture built for a classroom, not bolted onto one,” said Jian (Kevin) He, co-founder and CTO. “A district, school or teacher can give their AI Assistant its own skills and tools and benefit from the fine-tuning and RAG system that holds district documents, policies, curricula, and information. Those AI Assistants can have sub-agents of their own, making Colleague AI the first K-12 platform built to let schools design and deploy real multi-agent systems — and every one of them inherits the same FERPA, COPPA, GDPR and SOC 2 posture already documented in our data processing agreements.”

Additionally, a new knowledge graph gives the platform persistent memory of every teacher and student it serves. Its memories track learner progression, teacher professional growth, and organizational improvement. That memory is fundamental to improving the quality of AI outputs in an educational context: an assignment, a lesson or a piece of feedback shaped by what the platform already knows about a specific student — strengths, gaps, accommodations, pace — rather than something generated, generic, and not personalized. Further, the graph means AI output will get better the more it knows, instead of starting over with every request, which is exactly what all other competitors currently do. This is the power to truly enact personalized learning at scale, giving learning opportunities to all students without increasing teacher or administrative burdens.

For school and district clients, a new administrative dashboard gives administrators one view into every AI Assistant and sub-agent in use, who built it, and its usage. Districts are already building a school improvement plan assistant from their own data, a grading-consistency checker tuned to their own rubric language, and an IEP drafting tool shaped around their own documentation — tooling with no fixed feature set, and no general-purpose AI account, was built to offer.

For teachers, work that once meant moving between a lesson planner, a grading tool and a newsletter builder can now happen within one conversation, with each task informed by the context of a specific student or class. Colleague AI describes this as a shift toward 10× output, reducing repetitive setup while keeping teachers responsible for instructional judgment. One educator reported that adapting an entire unit into project-based learning took roughly 100 hours before using AI, then around 12 hours on average with AI. With Colleague AI’s new features, this specific workflow took 30 minutes. “I can spend my time on the stuff that actually needs me: Is this accurate? Is it rigorous? Will my kids care?”

For district leaders, the same approach supports building and maintaining tools in-house, including school improvement plans and intervention workflows tailored to local needs, rather than waiting for a vendor’s roadmap or an outside consultant.

The new product features will be released to enterprise clients later this month before being rolled out across the user base to paid and freemium users. Districts and schools interested in the Colleague AI platform with its governed and customized AI tools and Assistants can learn more at colleague.ai.

About Colleague AI
Colleague AI is the leading comprehensive, research-backed K-12 AI infrastructure platform serving teachers, students, and school leaders. The platform services the entire teaching and learning cycle, offering unlimited educator and administrative tools, easy-to-use embedded workflows, and AI management infrastructure. Built with more than $19 million in federal research funding from the NSF, U.S. Department of Education and Institute of Education Sciences, and developed in partnership with AmplifyLearn.AI at the University of Washington College of Education, Colleague AI serves 300K+ users nationwide and in 50+ countries. For more information, visit colleague.ai.

Naomi Thomas
Colleague AI
naomi.thomas@colleague.ai
Visit us on social media:
LinkedIn
Facebook
YouTube
X

Media gallery

About The Author

Written By: EIN Presswire