Edumate - a web app to connect tutor and tutee

🌟

For product demo please follow the links: Tutors match and Tutor's dashboard

1. Executive Summary & Core Value Proposition

Edumate is a highly structured, two-sided online marketplace designed to eliminate information asymmetry, manual search overhead, and trust deficits in the Singapore private tuition sector. The platform divides its user experience into two distinct UI modules that interact with a shared, highly secure backend data ecosystem.

For Parents/Tutees (The Discovery Engine): Replaces traditional, friction-heavy agency requests and static directory listings with a gamified, swipe-to-match UX pattern backed by an algorithmic pairing core.

For Tutors (The Operations Cockpit): Provides a comprehensive software-as-a-service (SaaS) ERP workspace to schedule sessions, view financial insights, record student telemetry, and utilize generative content-creation systems.

2. Monetization & Product Packaging Specifications

The platform relies on a differentiated freemium framework to monetize both sides of the marketplace without creating friction for early-stage adoption

2.1 The Parent "Student Pass" Tier

To optimize the onboarding funnel, parents are funneled through a strict feature-gated tier system:

Freemium Tier Limitations: Users who do not purchase the Student Pass are restricted to basic subject-based filtering, capped visibility on overall tutor profile details, and locked communications.

The Student Pass Token: A one-time commercial acquisition unlocking advanced multi-parameter filtration, algorithmic recommendation feeds, and in-app messaging.

Churn & Re-activation Gate: If a parent user goes digitally inactive for three consecutive months, the Student Pass token expires. Re-accessing premium pipelines requires a pass renewal purchase.

2.2 Tutor Premium Tools Workspace

Tutors onboard with zero commission charges on their core earnings to maximize supply acquisition. Monetization is driven by an auto-renewing monthly subscription giving access to advanced professional utilities:

Feature Set	Freemium Package	Paid Premium Package
Financial Cut	100% Earnings Retention	100% Earnings Retention
Profile Visibility	Standard Organic Feed Listing	Standard Organic Feed Listing
Core Utilities	Basic Profile Portfolio Page	Comprehensive Calendar Sync & Reminders
Analytics Core	None	Private Student Progress Logging Dashboard
AI Enhancements	None	Generative AI Personalized Lesson Material Engine

2.3 B2B Advertising Placement Engine

Tuition centers and educational brands buy targeted visual real estate through a standalone merchant console.

Core Auction Ad-Delivery: Powered by a Generalized Second-Price (GSP) auction model. Bidders submit maximum bids for targeted placements; winners pay the bid of the competitor directly below them plus a minimum increment ($0.01). This eliminates tactical bidding spirals common in first-price models.

Ad Targeting: Driven by a Contextual Multi-Armed Bandit (MAB) recommender system. It continuously balances exploitation (serving historically high-converting ads based on user cohort properties) and exploration (introducing fresh ad inventory to uncover new high-conversion profiles).

3. Product Module & Feature Specs

3.1 Parent Discovery Engine (`/find-tutor`)

Designed to solve user decision fatigue by replacing dense tabular text directories with an intuitive swipe interface.

1. Detailed Feature Breakdown

The Selection Swiper: Renders unified profile views displaying headshots, credentials, base pricing indices (Price range: $40-$60), and star-rating tags derived from historical data. Swiping denotes explicit intent to connect.

The Hard-Constraint Filter Sheet: A persistent overlay configuration menu enforcing programmatic restrictions on: Subject Category, Geographic Boundary (e.g., Serangoon), Weekly Budget, Session Duration, and Availability Window.

Virtual Pre-Screening Module: An interface allowing pass-holding parents to request and conduct live, short-form video interviews with prospective tutors before formally contracting a paid session.

2. Data Science & Algorithmic Mechanics

The backend pairs users through a multi-layered matching engine designed to bypass data collection gaps:

Overcoming Data Sparsity & Cold Starts: New parents enter via a mandatory structural questionnaire. The initial feed uses Content-Based Filtering to pair explicit parent needs with corresponding tutor backgrounds. As transactional density builds, the system blends into Collaborative Filtering, updating recommendation weights based on implicit platform interactions.

3.2 Tutor Workspace Dashboard (`/tutor-workspace`)

An administrative cockpit enabling tutors to view operational analytics and log student performance updates.

1. Core Component Matrix

The Financial Health Ledger: A high-level scorecard component array highlighting Total Earnings, Pending Accounts Receivable (escrow tracking), and total cumulative Sessions completed.

The Time-Series Demand Chart: An analytical line chart plotting historical revenue variations across continuous multi-month spans to assist instructors with business capacity planning.

Subject Distribution Graph: A bar chart visualizing absolute session volume broken down by subject vertical (e.g., Mathematics, Physics, Chemistry, Biology).

Student Telemetry Console ("Student Performance"): A granular time-series tracking chart displaying performance milestones over time. It includes dynamic filters to view metrics by a single student identifier (e.g., Alex Chen) or a localized subject scope.

2. Fraud Mitigation & Trust Protocol

To ensure platform integrity and eliminate profile substitution fraud, the workspace includes strict automated identity validation gates:

Onboarding Verification: Tutors must pass a mandatory credential review (academic qualification uploads) and a baseline facial enrollment check.

CNN Facial Verification Pipeline: Built on Convolutional Neural Networks (CNN), the application triggers randomized, periodic biometric authentication requests during normal tutor log-ins and before high-value session completions. Tutors must successfully clear the face match check against their enrollment vector to unlock their dashboard functionality.

4. Operational Success Metrics (Product KPI Dashboard)

To ensure long-term health across both nodes of the marketplace, engineering and product performance are tracked across four core categories.

4.1 North Star Metric (NSM)

Successful Match Conversion Rate:

Product Rationale: This single metric ensures alignment across both target users. It confirms that the matching algorithm is accurate, the onboarding process is low-friction, and both parties are exchanging real commercial value.

4.2 Secondary Performance Framework

4.3 Engineering SLA Targets

Algorithmic Match Latency: The Hungarian matching execution and vector search inferences must post a P95 latency response time under 350 milliseconds.

Facial Verification Throughput: CNN-based authentication requests must process image payloads and return match results within a P95 threshold of 1200 milliseconds on mobile networks.