AI-Powered Classroom Management at Scale

How Sijil streamlined student records, attendance, and grading with a 22-tool AI Academic Agent and Generative UI.

How Sijil streamlined student records, attendance, and grading with a 22-tool AI Academic Agent and Generative UI.

Document agent

Lead router

Client

Sijil · Classroom Management Platform

Industry

Team

Lead Full-Stack Engineer

Build

3 months (07/2026 – 09/2026)

Sheet

C-01

FIG. C-01 — Sijil Classroom Management & AI Agent Console

FIG. C-01 — Sijil Classroom Management & AI Agent Console

Scale 1 : 1

01

Survey — The problem

Managing classrooms required scattered spreadsheets and repetitive manual workflows.

Managing classrooms required scattered spreadsheets and repetitive manual workflows.

Schools and academic institutions often struggle with fragmented tools for student records, attendance tracking, and grading. Sijil solves this by unifying all academic workflows into an intelligent SaaS platform.

Architecture & Capabilities

  • AI Academic Agent: Equipped with 22 registered tools connected to the Sijil backend, capable of rendering interactive React UI components such as sortable data tables, student-selection forms, and confirmation flows instead of text-only answers.

  • Enterprise SaaS Foundation: Multi-tenant data isolation, rate limiting, token quotas, Docker-based CI/CD, and Cloudflare R2 presigned uploads.

  • Tech Stack: Next.js, React, NestJS, MongoDB, and Google Cloud Run using a mobile-first interface.

Exceptions / day

380

Handling time

6 min each

People on it

4 planners

02

Blueprint — What we built

One agent, three exits.

One agent, three exits.

01 · InputCarrier email + EDI380 / day
02 · AgentException agentRead · Match · Fix · Check
03a · OutputERP order fixedAuto · 81%
03b · OutputCustomer notifiedAuto · 81%
03c · ReviewPlanner approves> €500 · 19%
0.8 s Avg. decision
01 · InputCarrier email + EDI380 / day
02 · AgentException agentRead · Match · Fix · Check
03a · OutputERP order fixedAuto · 81%
03b · OutputCustomer notifiedAuto · 81%
03c · ReviewPlanner approves> €500 · 19%
0.8 s Avg. decision

03

Result — Measured after 90 days

−0%
−0%

exception handling time

0 h
0 h

returned to planners per day

0.0%
0.0%

accuracy on weekly audit

“Our planners stopped firefighting. The agent handles the boring 81%, and the 19% it hands over arrives already solved.”

“Our planners stopped firefighting. The agent handles the boring 81%, and the 19% it hands over arrives already solved.”

“Our planners stopped firefighting. The agent handles the boring 81%, and the 19% it hands over arrives already solved.”

Daan Visser

Head of Operations, Fieldwork Supply

Let’s build
something useful.

Have a software engineering opportunity or an AI application to discuss? Reach me at akilalsharafi@gmail.com. Based in Ibb, Yemen; working in Arabic and English.

Output · 01

Let’s build
something useful.

Have a software engineering opportunity or an AI application to discuss? Reach me at akilalsharafi@gmail.com. Based in Ibb, Yemen; working in Arabic and English.

Output · 01

Let’s build
something useful.

Have a software engineering opportunity or an AI application to discuss? Reach me at akilalsharafi@gmail.com. Based in Ibb, Yemen; working in Arabic and English.

Output · 01

Let’s build
something useful.

Have a software engineering opportunity or an AI application to discuss? Reach me at akilalsharafi@gmail.com. Based in Ibb, Yemen; working in Arabic and English.

Output · 01