Education · End-to-End AI
End-to-end AI implementation at NookTek
From an AI roadmap on a whiteboard to learning copilots in students' hands — the data platform, the AI features, and the team to run them, delivered as one engagement.
The challenge
NookTek had an ambitious AI roadmap and real competitive pressure — but student data scattered across their LMS, CRM, and product database, no ML infrastructure, and an engineering team fully booked on the core product. Every AI feature was six months away, permanently.
What we built
A four-person AcePointe team embedded with NookTek's engineers and shipped in layers:
- Unified data platform — student activity, outcomes, and content brought into one governed warehouse with privacy controls appropriate for minors' data.
- Learning copilot — an in-product tutor that explains concepts against the student's own coursework, with guardrails reviewed by NookTek's pedagogy team.
- Educator analytics — early-warning dashboards flagging students who are struggling weeks before grades show it.
- In-house capability — paired delivery throughout; NookTek engineers co-owned every service from the first commit.
The outcome
The copilot shipped to a pilot cohort in week six and to all users within the quarter. Session length and course completion both moved. When we rolled off, NookTek's own team shipped the next two AI features without us — which is exactly how we measure success.