From what to build to how to build it — our platform helps you understand, plan, and solve SIH 2026 problem statements with complete insights into requirements, solution approaches, technology stacks, architecture, AI/ML, and implementation roadmaps.
Search for your problem statement, and start building your solution.
| Description: | The solution should include secure signup and login functionality with three user roles: Trainee, Trainer, and Admin. Trainees should be able to create professional profiles with qualifications, work experience, interests, skills, and certificates, enroll in courses, access learning resources, attempt subject-wise MCQ assessments, and provide feedback on courses and training content.Trainers should be able to manage their profiles, create questionnaires with deadlines, monitor trainee participation and performance, and upload recorded lectures, presentations, and study materials in a trainer library accessible to trainees.The Admin module should provide user approval and role management features along with dashboards for monitoring courses, enrollments, certifications,assessments, and participation statistics. Admins should also be able to publish notifications, announcements, achievements, and newly added learning content on the homepage.The platform should support competency mapping for identifying suitable trainers for various subjects and should be scalable, secure, user-friendly, and accessible across devices to promote efficient learning and organizational capacity building. |
| Organization: | Ministry of Earth Sciences (MoES) |
| Department: | India Meteorological Department |
| Category: | Software |
| Theme: | Smart Educationtd> |
| Youtube Link: | Not provided |
| Dataset Link: | Not provided |
| Contact info: | Not provided |
CAPACITY CONNECT is a centralized, web-based Digital Capacity Building and Learning Management Portal designed to standardize, manage, and scale specialized scientific and technical training across the Ministry of Earth Sciences (MoES) and the India Meteorological Department (IMD) [PS]. The platform integrates role-based workflows for Trainees, Trainers, and Admins, offering comprehensive profile tracking, course delivery, media library management, deadline-enforced MCQ evaluations, and admin analytics [PS]. Beyond standard LMS capabilities, CAPACITY CONNECT solves the critical organizational challenge of trainer discovery by incorporating an automated competency mapping engine that pairs course subject requirements with qualified trainer profiles based on verified qualifications, experience, and domain expertise [PS, INFERENCE]. Built with a modern responsive frontend, scalable API backend, and offline-resilient assessment engine, it enables continuous skill development across both central offices and bandwidth-constrained remote meteorological observatories [INFERENCE].
Fundamental System Objective:
To establish a secure, centralized, role-based learning management and competency mapping platform that streamlines technical training, automates assessment workflows, and dynamically pairs qualified domain experts with training programs across MoES/IMD institutes [PS, INFERENCE].
The problem description from the official SIH specification has been broken down into atomic requirements below, maintaining complete fidelity to the original text:
REQ-01: Authentication & Access Control
/api/trainee/*, /api/trainer/*, /api/admin/*) [INFERENCE]. REQ-02: Trainee Profile Management
REQ-03: Course Enrollment & Resource Access
trainee_id, course_id, progress_pct, status) and resource delivery endpoint supporting video streaming and PDF viewing [INFERENCE]. REQ-04: Subject-Wise MCQ Assessments
REQ-05: Feedback Engine
REQ-06: Trainer Profile Management
REQ-07: Questionnaire & Deadline Management
REQ-08: Participation & Performance Monitoring
REQ-09: Trainer Resource Library Uploads
REQ-10: Admin User Approval & Role Management
pending, approved, rejected), role assignment, and privilege revocation [INFERENCE]. REQ-11: Admin Monitoring Dashboards
REQ-12: Admin Homepage Content Management (CMS)
REQ-13: Competency Mapping & Trainer Matching
REQ-14: Cross-Device Accessibility, Security, and Scalability
| Stakeholder | Goal | Current Pain | System Need |
|---|---|---|---|
| IMD Trainees (Meteorological Officers, Scientists, Field Observers) | Access domain training, complete certifications, and log career skill progression. [INFERENCE] | Fragmented physical offline training, lack of centralized record for certifications, missed deadlines due to field postings. [RESEARCHED] | Unified portal to view assigned courses, stream video lectures, take timed MCQs, track certified skills, and download course certificates. [PS, INFERENCE] |
| Trainers & Subject Matter Experts (Senior Scientists at MTI Pune, ICITC Delhi) | Deliver structured course content, evaluate officer knowledge, and maintain teaching materials. [RESEARCHED] | Manual evaluation of physical paper tests, unstructured email sharing of presentation slides, no clear tracking of trainee progress. [RESEARCHED] | Trainer studio for managing slide uploads, uploading video lectures, building automated dynamic quizzes with deadlines, and viewing score distributions. [PS, INFERENCE] |
| MoES / IMD HR & Training Admins (Training Cell Coordinators) | Manage user access, monitor national training statistics, publish announcements, and allocate domain trainers. [RESEARCHED] | Inability to track department-wide training completion, manual search for qualified trainers across institutes, unverified account creations. [RESEARCHED] | Account approval workflow, system-wide analytics dashboard, homepage CMS, and automated competency mapping engine to search/assign trainers. [PS, INFERENCE] |
| Executive Leadership (Director General of Meteorology, MoES Capacity Division) | Oversee organizational capacity building, ensure compliance with national training policies (Mission Karmayogi). [RESEARCHED] | Lack of consolidated real-time reporting on technical capacity gaps across coastal, radar, and forecasting units. [RESEARCHED] | Executive KPI dashboard summarizing total certified officers, active enrollments, subject skill coverage, and training feedback trends. [PS, INFERENCE] |
| Proposed Feature Module | Target Requirement ID | Technical Component | Justification |
|---|---|---|---|
| Role-Based Auth & Guarding | REQ-01, REQ-10 | JWT + Argon2 ID Hashing + RBAC Middleware | Ensures secure isolation between Trainee, Trainer, and Admin capabilities [PS]. |
| Trainee Profile & e-Portfolio | REQ-02 | Profile Schema + S3 Certificate Storage | Stores qualifications, experience, and uploaded credentials for skill mapping [PS]. |
| Course Catalog & Player | REQ-03, REQ-09 | React/Next.js Player + HLS Video Streaming | Delivers recorded lectures, slides, and study notes seamlessly across devices [PS, INFERENCE]. |
| Timed MCQ Assessment Engine | REQ-04, REQ-07 | Dynamic Quiz Engine + Auto-grader | Enforces quiz submission deadlines, auto-evaluates MCQs, and logs attempt records [PS]. |
| Feedback & Rating System | REQ-05 | Feedback Form + Aggregation Pipeline | Captures trainee satisfaction scores to guide trainer material updates [PS]. |
| Trainer Content Studio | REQ-06, REQ-09 | Trainer Dashboard + S3 Media Uploader | Empowers trainers to maintain study libraries and manage course offerings [PS]. |
| Admin User Governance | REQ-10 | Approval Queue UI + Role Modifier | Restricts system access to verified MoES/IMD personnel [PS]. |
| Executive Analytics Dashboard | REQ-11, REQ-08 | Recharts Visualizations + SQL Metrics API | Provides real-time visibility into course participation, test scores, and compliance [PS]. |
| Homepage CMS Engine | REQ-12 | CMS Publishing API + Banner Carousel | Enables admins to broadcast urgent announcements and feature top achievements [PS]. |
| Competency Mapping Engine | REQ-13 | Vector Cosine Similarity / TF-IDF Matcher | Automates candidate trainer search by indexing skill requirements against profiles [PS, INFERENCE]. |
| Responsive PWA Shell | REQ-14 | TailwindCSS + Service Workers | Ensures offline assessment tolerance and cross-device mobile support [PS, INFERENCE]. |
| Data Type | Purpose | Ideal Real-World Source | Hackathon Substitute | Access Method | Availability | Privacy Sensitivity | Confidence |
|---|---|---|---|---|---|---|---|
| User Profiles & Qualifications | Authenticate users and establish trainee/trainer skill vectors. | IMD HRMS / SPARROW Portal database [RESEARCHED]. | Synthetic JSON dataset of 100 MoES profiles with domain skills. | Seeded SQL / REST API | High (Synthetic) | High (Contains PII) | [ASSUMPTION] |
| Competency & Skill Taxonomy | Standardize competency tags for mapping trainers to subjects. | Mission Karmayogi FRAC / KCM Competency Dictionary [RESEARCHED]. | Structured JSON array based on MTI Pune syllabus disciplines [RESEARCHED]. | Static File / DB Table | High | Public / Low | [RESEARCHED] |
| Course Media & Slides | Deliver educational lectures, PDFs, and presentations. | MTI Pune e-Learning archive & lecture repositories [RESEARCHED]. | Open-access MoES scientific PDFs & sample MP4 video lectures. | Local Storage / S3 Bucket | High | Public / Low | [RESEARCHED] |
| Questionnaires & Question Banks | Test subject knowledge through scheduled MCQs. | MTI Pune Question Bank archives [RESEARCHED]. | Curated JSON file containing 50 subject-categorized MCQs. | REST API / PostgreSQL | High | Medium | [RESEARCHED] |
| Enrollments & Test Results | Log progress, track metrics, and generate performance analytics. | Production LMS database logs. | Simulated execution logs generated during test runs. | PostgreSQL DB queries | High | Medium | [INFERENCE] |
| Homepage CMS & Banners | Display announcements, news, and achievements on the landing page. | MoES / IMD Press Information Bureau releases [RESEARCHED]. | Mock news JSON objects featuring recent MoES scientific achievements. | CMS REST API | High | Public / Low | [RESEARCHED] |
iGOT Karmayogi Platform (Government of India)
MTI Pune Web Portal (Existing IMD Training Site)
Generic Open-Source LMS (Moodle / Canvas LMS)
+---------------------------------------------------------------------------------------------------+
| CANDIDATE ARCHITECTURAL APPROACHES |
+------------------------------------+--------------------------------+-----------------------------+
| Approach 1: Monolithic Moodle | Approach 2: Full-Stack React + | Approach 3: Fully Serverless|
| Customization (PHP Stack) | Node/PostgreSQL + Hybrid Match | + LLM-Driven Adaptive LMS |
+------------------------------------+--------------------------------+-----------------------------+
| • Legacy PHP/MySQL core | • React/Next.js modern UI | • AWS Lambda + DynamoDB |
| • Custom Moodle PHP Plugins | • Express/PostgreSQL backend | • OpenAI/Claude LLM API |
| • Heavily plugin dependent | • Dedicated Vector Match Engine| • Fully dynamic AI quizzes |
| • Monolithic & hard to audit | • Modular, lightweight, fast | • Unpredictable cloud cost |
+------------------------------------+--------------------------------+-----------------------------+
| Evaluation Dimension | Approach 1: Monolithic Moodle Customization | Approach 2: Modern Full-Stack (Node/React + PostgreSQL + Vector Engine) | Approach 3: Fully Serverless + LLM-Driven Adaptive LMS |
|---|---|---|---|
| Problem Fit | Medium (Standard LMS capabilities, weak custom matching) [INFERENCE] | High (Custom-tailored to explicit PS requirements) [INFERENCE] | Medium (Over-engineered, AI hallucination risk in MCQs) [INFERENCE] |
| Accuracy Potential | Medium (Rule-based plugins) [INFERENCE] | High (Deterministic vector cosine scoring for competency) [INFERENCE] | Unpredictable (Non-deterministic LLM output) [INFERENCE] |
| Data Requirements | Standard SQL Schema [KNOWN] | Relational SQL + Structured Skill Taxonomy vectors [INFERENCE] | Large prompt contexts + Cloud API tokens [INFERENCE] |
| Complexity | High (Navigating legacy PHP codebases) [INFERENCE] | Moderate (Clean separation of concerns) [INFERENCE] | Very High (Distributed serverless debugging) [INFERENCE] |
| Explainability | High [KNOWN] | High (Mathematical similarity match scores) [INFERENCE] | Low (Black-box LLM scoring) [INFERENCE] |
| Cost Profile | Low (Open source) [KNOWN] | Low ($0 for hackathon; standard VM hosting) [INFERENCE] | High (Continuous LLM API call consumption) [INFERENCE] |
| Scalability | Moderate (Requires vertical server scaling) [INFERENCE] | High (Stateless API containers + indexed DB) [INFERENCE] | High (Serverless auto-scaling) [KNOWN] |
| Hackathon Feasibility | Low (Difficult to demo clean modern UI) [INFERENCE] | Very High (Fully buildable by 4-6 students in window) [ASSUMPTION] | Moderate (Dependency on external API reliability) [INFERENCE] |
| Production Feasibility | Moderate [KNOWN] | High (Easy to host on NIC/MeghRaj cloud) [INFERENCE] | Low (Government cloud security/cost constraints) [INFERENCE] |
Selected Strategy: Approach 2 — Modern Full-Stack Architecture with a Dedicated Competency Matching Engine [ASSUMPTION].
Approach 2 achieves the optimal balance between rapid hackathon execution and production-grade engineering [INFERENCE]. By utilizing a React/Next.js responsive frontend connected to a RESTful Node.js/Express API backed by PostgreSQL, the team can deliver a slick, responsive user experience across all three user personas [INFERENCE]. The competency mapping requirement is solved using an explainable Vector Cosine Similarity algorithm that compares subject skill vectors against trainer competency vectors, augmented by experience weighting [INFERENCE].
+---------------------------------------------------------------------------------------------------+
| CAPACITY CONNECT ARCHITECTURE |
+---------------------------------------------------------------------------------------------------+
[ Trainee Browser / Mobile ] [ Trainer Browser / Desktop ] [ Admin Browser / Dashboard ]
| | |
+------------------------------------+-----------------------------------+
|
v
+---------------------------------------------------------------------------------------------------+
| CLIENT / FRONTEND LAYER (PWA) |
| • Next.js 14 / React (App Router) • TailwindCSS & Shadcn UI • Service Worker (Offline Cache) |
| • HTML5 Video Player (HLS) • Formik / Zod Validation • Lucide / Recharts Analytics UI |
+---------------------------------------------------------------------------------------------------+
| (HTTPS / REST APIs / JWT Auth)
v
+---------------------------------------------------------------------------------------------------+
| APPLICATION / BACKEND LAYER |
| • Express.js REST API Core • JWT / Argon2 Auth Guard • Multer Media Upload Middleware |
| • Competency Engine (TF-IDF/Vector) • Quiz Evaluation Subsystem • CMS Content Publisher |
+---------------------------------------------------------------------------------------------------+
|
+-------------------------+-------------------------+
| |
v v
+-----------------------------------------------+ +-----------------------------------------------+
| DATABASE LAYER | | MEDIA & STORAGE LAYER |
| • PostgreSQL 16 (Relational Core) | | • MinIO / S3 Object Storage |
| • JSONB (Skill vectors & Quiz Schemas) | | • Lectures (MP4/HLS), Slides (PDF/PPTX) |
| • Indexed User, Course & Result tables | | • User Qualification Certificates |
+-----------------------------------------------+ +-----------------------------------------------+
User Authentication & Role Activation:
Trainee or Trainer) [PS]. is_approved = false in PostgreSQL [INFERENCE]. user_id, email, and role [INFERENCE].Trainer Course & Quiz Setup Flow:
Trainee Learning & Assessment Flow:
assessment_attempts table, updating user progress metrics [INFERENCE]. Competency Mapping Flow:
| Layer | Selected Technology | Why Selected | Alternative Considered | Why Alternative Rejected |
|---|---|---|---|---|
| Frontend Framework | Next.js 14 / React | Server-side rendering for fast initial loads, built-in routing, and easy PWA integration [INFERENCE]. | Pure HTML/jQuery | Lacks component reusability, complex state management for quizzes [INFERENCE]. |
| Styling & UI Components | TailwindCSS + Shadcn UI | Rapid development of responsive, accessible, government-standard user interfaces [INFERENCE]. | Bootstrap 5 | Heavy CSS bundle size, generic non-modern appearance [INFERENCE]. |
| Backend Runtime | Node.js + Express.js | Lightweight, high-throughput asynchronous I/O ideal for handling API routes and media metadata [INFERENCE]. | Python Django | Higher overhead for simple REST APIs during a hackathon [INFERENCE]. |
| Database | PostgreSQL 16 | Robust relational compliance, native JSONB support for quiz structures, and strong index performance [INFERENCE]. | MongoDB | Weaker relation handling across courses, users, and enrollments [INFERENCE]. |
| Competency Engine | Python (scikit-learn) / Node-vector | Highly efficient TF-IDF vector processing and cosine similarity calculations for trainer matching [INFERENCE]. | Complex Graph DB (Neo4j) | Overkill for hackathon scale; introduces unnecessary infrastructure overhead [INFERENCE]. |
| Media Storage | MinIO (S3 API Compatible) | Self-hosted object storage compatible with AWS S3 APIs; ideal for local development and on-prem deployment [INFERENCE]. | Direct Local Filesystem | Harder to scale horizontally across multiple web app instances [INFERENCE]. |
| Authentication | JSON Web Tokens (JWT) + Argon2 | Stateless, secure, industry-standard authentication mechanism [INFERENCE]. | Session Cookies | Requires server-state memory, harder to scale across microservices [INFERENCE]. |
| Containerization | Docker & Docker Compose | Single-command deployment orchestrating DB, Backend, MinIO, and Frontend for judges [INFERENCE]. | Manual VM Install | Error-prone environment setup during judge evaluations [INFERENCE]. |
+---------------------------------------------------------------------------------------------------+
| COMPETENCY MAPPING ALGORITHM WORKFLOW |
+---------------------------------------------------------------------------------------------------+
[ Subject Requirement ] [ All Trainer Profiles ]
Skill Tags: ["Radar", " Doppler", Trainer A: ["Radar", "Satellite", Exp: 8 yrs, Rating: 4.8]
"Signal Processing"] Trainer B: ["Synoptic", "Radar", Exp: 3 yrs, Rating: 4.0]
| |
v v
[ Subject Skill Vector (S) ] [ Trainer Skill Vector Matrix (T) ]
| |
+---------------------------+---------------------------+
|
v
[ Cosine Similarity Scoring ]
Sim(S, T_i) = (S · T_i) / (||S|| * ||T_i||)
|
v
[ Weighted Final Ranking Score ]
Final Score = (0.6 * Sim) + (0.25 * Norm_Exp) + (0.15 * Avg_Rating)
|
v
[ Output: Ranked Trainer Recommendation List ]
Explicit Exclusion: Generative AI MUST NOT be used for automated MCQ grading or user role approvals to maintain 100% deterministic accuracy and governance compliance [INFERENCE].
-- Users Table (Handles all 3 roles)
CREATE TABLE users (
id SERIAL PRIMARY KEY,
full_name VARCHAR(150) NOT NULL,
email VARCHAR(150) UNIQUE NOT NULL,
password_hash VARCHAR(255) NOT NULL,
role VARCHAR(20) CHECK (role IN ('TRAINEE', 'TRAINER', 'ADMIN')) NOT NULL,
is_approved BOOLEAN DEFAULT FALSE,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
-- Trainee Profiles
CREATE TABLE trainee_profiles (
id SERIAL PRIMARY KEY,
user_id INT REFERENCES users(id) ON DELETE CASCADE,
qualifications TEXT[],
work_experience_years INT DEFAULT 0,
interests TEXT[],
skills TEXT[],
certificates_json JSONB DEFAULT '[]'::jsonb
);
-- Trainer Profiles
CREATE TABLE trainer_profiles (
id SERIAL PRIMARY KEY,
user_id INT REFERENCES users(id) ON DELETE CASCADE,
designation VARCHAR(100),
department VARCHAR(100),
expertise_skills TEXT[],
years_experience INT DEFAULT 0,
average_rating NUMERIC(3,2) DEFAULT 0.00
);
-- Courses Table
CREATE TABLE courses (
id SERIAL PRIMARY KEY,
title VARCHAR(200) NOT NULL,
description TEXT,
category VARCHAR(100),
created_by INT REFERENCES users(id),
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
-- Questionnaires / Assessments
CREATE TABLE assessments (
id SERIAL PRIMARY KEY,
course_id INT REFERENCES courses(id) ON DELETE CASCADE,
title VARCHAR(200) NOT NULL,
deadline TIMESTAMP NOT NULL,
pass_mark INT DEFAULT 50,
questions_json JSONB NOT NULL -- Array of {q_id, question, options[], correct_idx}
);
-- Assessment Attempts
CREATE TABLE assessment_attempts (
id SERIAL PRIMARY KEY,
assessment_id INT REFERENCES assessments(id) ON DELETE CASCADE,
user_id INT REFERENCES users(id) ON DELETE CASCADE,
score INT NOT NULL,
passed BOOLEAN NOT NULL,
submitted_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
-- Course Feedback
CREATE TABLE course_feedback (
id SERIAL PRIMARY KEY,
course_id INT REFERENCES courses(id) ON DELETE CASCADE,
user_id INT REFERENCES users(id) ON DELETE CASCADE,
rating INT CHECK (rating BETWEEN 1 AND 5),
comments TEXT,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
localStorage. If power or network drops, answers remain intact upon browser refresh [INFERENCE]./api/trainee/submit-quiz [INFERENCE]./api/trainer/*) or admin queues (/api/admin/*) [INFERENCE].is_approved = false), blocking system access until verified by an Administrator, preventing unauthorized external access [PS, INFERENCE].audit_logs database table [INFERENCE].+---------------------------------------------------------------------------------------------------+
| HACKATHON MVP SCOPE BOUNDARY |
+---------------------------------------------------------------------------------------------------+
[ IN SCOPE (MUST BUILD FOR SIH) ]
+-------------------------------------------------------------------------------+
| • Auth & Role Approval Workflow (Trainee, Trainer, Admin) |
| • Trainee Profile Builder (Skills, Qualifications, PDF Uploads) |
| • Trainer Media Library Uploader (Lectures, Slides, Documents) |
| • Deadline-Enforced MCQ Assessment Engine with Auto-Grading |
| • Admin Monitoring Dashboard & Homepage CMS Announcement Engine |
| • Vector Cosine Similarity Competency Engine for Trainer-Subject Matching |
+-------------------------------------------------------------------------------+
|
v
[ OUT OF SCOPE / DESCRIBED ONLY (FUTURE) ]
+-------------------------------------------------------------------------------+
| • Live WebRTC Video Classroom Integration (Jitsi / Zoom Embed) |
| • DigiLocker API Integration for External Certificate Verification |
| • Native iOS / Android Mobile Apps (PWA web shell used instead) |
| • SCORM / xAPI LMS Enterprise Package Import/Export Engine |
+-------------------------------------------------------------------------------+
To deliver a compelling, flawless demo during SIH judging without external API dependencies, the system uses pre-seeded, realistic MoES/IMD scientific data [INFERENCE]:
Step 1: Admin Persona (2 Mins)
[email protected]. Step 2: Trainer Persona (2 Mins)
[email protected]). Step 3: Trainee Persona (3 Mins)
[email protected]). Step 4: Competency Mapping Showcase (2 Mins)
Target Deadline: 30 September 2026 [PS]
+---------------------------------------------------------------------------------------------------+
| SIH IMPLEMENTATION ROADMAP |
+---------------------------------------------------------------------------------------------------+
Phase 1: Core Arch & Auth Setup [ Weeks 1 - 2 ] ═══════════════════════╗
• DB Schema & Docker Setup ║
• JWT Auth & Admin Approval Queue ║
v
Phase 2: Trainee & Trainer Features [ Weeks 3 - 4 ] ═══════════════════════╗
• Profiles, Skill Vectors & Library Uploads ║
• Course Enrollment & Media Player ║
v
Phase 3: Assessments & CMS [ Weeks 5 - 6 ] ═══════════════════════╗
• Deadline MCQ Engine & Auto-Grading ║
• Admin CMS Banners & Feedback Pipeline ║
v
Phase 4: Competency Mapping Engine [ Weeks 7 - 8 ] ═══════════════════════╗
• TF-IDF Vectorization Algorithm ║
• Admin Trainer Matching & Scoring UI ║
v
Phase 5: Hardening & Demo Preparation [ Weeks 9 - 10 ] ═══════════════════════╝
• Seed MoES Data, PWA Cache, Stress Testing & Final Demo Dry Run (Sept 30, 2026)
| Test Category | Target Component | Test Methodology | Success Criteria |
|---|---|---|---|
| Functional | Auth, Profile, Quiz, CMS | Automated E2E testing using Playwright scripts. | 100% pass on all CRUD and role boundary checks [INFERENCE]. |
| Performance | API & Database | Load testing using k6 simulating 200 concurrent users. |
Sub-200ms API response time, zero DB connection timeouts [INFERENCE]. |
| Usability | Cross-Device UI | Testing across desktop, tablet (768px), and mobile (360px) viewports. | Responsive rendering with WCAG 2.1 AA contrast compliance [INFERENCE]. |
| Competency Matching | Vector Engine | Precision@K benchmark against 20 hand-labeled trainer match cases. | Greater than 85% top-3 trainer match accuracy [INFERENCE]. |
| False Positive/Negative | User Approvals & Quizzes | Verify non-approved users cannot hit API endpoints; verify incorrect quiz answers fail. | Zero permission leakages or grading errors [INFERENCE]. |
| Security | OWASP Top 10 | Automated vulnerability scanning using OWASP ZAP. | Zero critical or high-risk vulnerabilities found [INFERENCE]. |
| Offline Resilience | MCQ Quiz Player | Simulating network disconnection during active quiz attempt. | Client state auto-saved in local storage; successfully synced upon reconnect [INFERENCE]. |
| Risk Event | Likelihood | Impact | Severity | Mitigation Strategy |
|---|---|---|---|---|
| Large File Video Upload Failures | High | Medium | High | Implement chunked multi-part upload via Multer to object storage with progress indicators [INFERENCE]. |
| Unapproved Account Data Leakage | Medium | High | High | Strict RBAC middleware checking is_approved === true on all operational endpoints [INFERENCE]. |
| Competency Engine Keyword Mismatch | Medium | Medium | Medium | Normalize input terms using standardized skill taxonomies (KCM framework) before vector conversion [INFERENCE]. |
| Network Loss During Exam Submission | High | High | High | LocalStorage state persistence + Background Sync API for automatic retry upon connection recovery [INFERENCE]. |
| Hackathon Scope Creep | High | Medium | High | Freeze MVP features at Phase 3; focus strictly on PS explicit requirements [ASSUMPTION]. |
| Key Capability | Generic LMS Submission | CAPACITY CONNECT Proposed Solution | Requirement Alignment |
|---|---|---|---|
| Competency Mapping | Static manual tag search or absent. | Vector Cosine Similarity engine calculating match scores between subject needs and trainer profiles [INFERENCE]. | Direct fulfillment of REQ-13 [PS]. |
| Domain Taxonomies | Generic academic subjects (Math, Science). | Pre-loaded with MoES / IMD domain disciplines (Radar, Synoptic, Hydrometeorology, Satellite) [RESEARCHED]. | Aligned with MoES/IMD organizational mandate [PS]. |
| Governance Workflow | Self-registration without verification. | Multi-tier admin account approval queue restricting access to verified staff [INFERENCE]. | Direct fulfillment of REQ-10 [PS]. |
| Field Resiliency | Requires stable continuous broadband. | Offline-tolerant PWA quiz player auto-saving state to IndexedDB/LocalStorage [INFERENCE]. | Fulfillment of REQ-14 [PS, INFERENCE]. |
[ASSUMPTION] PostgreSQL with vector extension / scikit-learn cosine similarity is sufficient for matching up to 5,000 trainer profiles efficiently. [ASSUMPTION] Hackathon judges will evaluate system capabilities using pre-seeded synthetic MoES/IMD user datasets. [NEEDS VALIDATION] Exact video transcode profile preferences for MoES internal IT network deployments. [NEEDS VALIDATION] Specific SSO integration mandates (e.g., Parichay API availability) for production rollout.Proceed immediately with Approach 2 (Modern Full-Stack React/Next.js + Express/PostgreSQL + Vector Competency Matching Engine) [ASSUMPTION]. This architecture fulfills 100% of the explicit requirements in Problem Statement SIH26075 while remaining realistic for a student team to complete within the hackathon timeframe [INFERENCE].
Availability of structured skill taxonomies and course categories representative of IMD/MoES training programs (derived from MTI Pune disciplines) [RESEARCHED].
Handling video lecture storage and streaming performance without degrading application response times during judge evaluations [INFERENCE].
Implementing an explainable, deterministic vector similarity algorithm for trainer competency mapping that produces accurate match rankings without relying on opaque black-box LLMs [INFERENCE].
users, trainee_profiles, trainer_profiles, courses, assessments, and feedback [INFERENCE]. is_approved) [INFERENCE]. docker-compose up) [INFERENCE].92 / 100 (Extremely high feasibility; core requirements rely on standard web paradigms, SQL database relations, and straightforward vector mathematics) [INFERENCE].
HIGH (The solution design strictly preserves all PS constraints, uses proven production technologies, and avoids fragile external dependencies) [INFERENCE].
Capacity Building Commission (CBC) - Karmayogi Competency Model (KCM)
Meteorological Training Institute (MTI), IMD Pune - Academics & Syllabi
Standard Operating Procedure for Training Activities in IMD
iGOT Karmayogi Digital Platform Portal
What we think about this platform, and how it can help people solve SIH problems.
" 30M+ Google Gemini tokens went into researching, analyzing, and understanding all 233 SIH problem statements to make them easier to explore and solve. "
" The platform makes SIH problem statements much easier to understand. Having everything from the problem details to the procedure in one place. "
" Oh, you got ads? Well, we need to make money too! But, you can use our platform to help you win SIH and take home the prize money. "
Get in touch if you need help bringing your ideas to life, or simply have an interesting idea you'd like to talk through. I'm always open to a conversation and exploring new possibilities together.