AI Olympiad
National AI Olympiad platform for Grades 5–12 — school portals, registrations, results, certificates, and city-scale operations. Built as a season operating system, not a brochure microsite.
Season operating flow
Every screen maps to a state in the live competition pipeline.
- Register School & student intake
- School portal Ops access for institutions
- Compete Junior / Senior pathways
- Score Season results ingestion
- Rank Deterministic leaderboards
- Certify Bulk certificate images
The challenge
A national AI olympiad fails when it is treated like a marketing site. Schools need portals, students need fair Junior (Grades 5–8) and Senior (Grades 9–12) pathways, admins need results and certificates, and parents need credibility in one place.
Fragmented registration, ranking, and certificate tools made every season brittle. A new centre, syllabus update, or results window meant manual cleanup and support tickets instead of product flows.
Tier-2 and Tier-3 school networks needed mobile-fast pages and self-serve answers — principals should not depend on WhatsApp threads to understand rules, centres, or rankers.
What we built
We engineered AI Olympiad as a multi-surface operations platform: public journeys (About, Journey, Rankers, Syllabus), school registration, student participation, admin tooling, results management, and certificate image generation.
Seasons, cities, schools, categories, and rankers are first-class data — not blog posts. That single decision is what lets analytics and AI layers plug in later without a rewrite.
Auth patterns (including email login flows), SEO architecture, and documented results workflows were shipped as product surfaces so each season reuses the same operating model.
Platform architecture
Four surfaces, one competition graph.
Public web
About, Journey, Rankers, Syllabus, SEO discovery for schools and parents.
School portals
Institution login, registration management, and season participation controls.
Admin & results
Results management, email login hardening, operational docs for season teams.
Certificates
Image generation pipeline that turns ranked outcomes into verifiable credentials.
Competition system diagram
Public, school, admin, and certificate surfaces sharing one season graph.
Public
Season
Truth
Ranking stays deterministic — AI assists messaging and insight, never winners.
Design principles
Product rules that kept the olympiad operable under real season pressure.
State machine first
UI maps to register → portal → compete → score → rank → certify — not decorative pages.
Deterministic ranking
AI can assist messaging and insight; it never invents a winner.
Self-serve content
About, Journey, Rankers, and Syllabus cut support load before tickets open.
Mobile-real performance
Budgets assume school Wi-Fi and mid-range Android — speed is acquisition.
How we approached it
We started from the season state machine, not the homepage: register → school portal → compete → score → rank → certify. Every UI decision had to serve one of those states.
Public content was designed to cut support load. About, Journey, Rankers, and Syllabus pages answer principal and parent questions before a ticket is raised.
Admin and results tooling got the same craft as the marketing site. Results Management docs and email-login hardening mean ops teams are not rediscovering the playbook every season.
Performance budgets assumed school Wi-Fi and mid-range Android phones. Discovery SEO and mobile speed were treated as acquisition channels for school participation.
We kept ranking deterministic and auditable. AI can assist messaging, insights, and future practice — it never invents a winner.
Where AI enters — and where it must not
The Olympiad product teaches AI literacy; the platform stores the competition graph that makes AI features honest. Student, school, city, category, score, and certificate events are relational truths.
That graph unlocks school-support assistants, registration anomaly detection, participation heatmaps, and personalized practice recommendations without rebuilding core tables.
Certificate and results pipelines stay rule-based for authenticity. Models can draft communications and surface insights; they cannot mutate official ranks.
As seasons compound, the same dataset becomes an intelligence layer: which cities convert portals, which schools stall mid-registration, which syllabus topics correlate with ranker strength.
Technical notes from production
Details that only show up once the season is real.
- Category rules are product logic: Junior (Grades 5–8) and Senior (Grades 9–12) compete within peer bands so fairness is enforced in software, not only in PDFs.
- School portal adoption is tracked as an ops signal — the platform surfaces registration gaps without exposing raw school rosters publicly.
- Certificate generation is treated as a batch systems problem: inputs are ranked records; outputs are branded images with auditability.
- Season docs (results, email login, SEO) ship with the codebase so handoffs do not depend on tribal knowledge.
Delivery & systems
Shipped as a PHP application with admin modules, assets pipeline, certificate image generation, Firebase-aligned auth patterns, and season copy systems for AIO messaging.
Internal registration and campaign surfaces keep season momentum visible for ops — without publishing school or student counts on the marketing site.
Documentation (results management, email login, SEO) travels with the code so season handoffs do not depend on tribal knowledge.
Stack
Click a layer — each one maps to a real season responsibility.
Application core
PHP
Server-rendered season site, admin modules, registration flows, and certificate generation hooks.
Brand assets from the live product
Primary mark pulled from the Olympiad codebase — shown once, on the surface it was built for.
Outcomes
- Unified public + school + admin olympiad stack
- Documented results & certificate season workflows
- Competition graph ready for analytics and AI assistants
- Junior/Senior pathways encoded in product, not PDFs alone
- Season handoffs via code-backed ops docs, not tribal knowledge
More work
View all
2024–2026
Archit Art Gallery
Commerce · Invoicing · Amazon AI Listings
Archit Art Gallery
Commerce · Invoicing · Amazon AI Listings
2023–2026
QUEST Book
EdTech · AI & STEAM Quiz
QUEST Book
EdTech · AI & STEAM Quiz
2024–2025
Campus Coach
EdTech · Counselling Platform
Campus Coach
EdTech · Counselling Platform