SMT / Selected engineering work

Case studies

How each project works, the decisions behind it, and the source code that supports the write-up.

  1. 01 / Institutional management system

    PGC-DHA

    Role-based access and staged admissions workflows
  2. 02 / AI-powered career toolkit

    ResumeIQ

    Semantic matching, evidence in experience, and PDF generation
  3. 03 / VS Code developer tool

    Ledger

    Execution capture, SHA-256 snapshots, and run comparison
  4. 04 / Offline document question answering

    Folio

    Local retrieval, page provenance, and source-quote checks
  5. 05 / React Native reminder application

    Nudge

    Reminder input, local persistence, and notification scheduling
Walk through PGC-DHA’s architecture

PGC-DHA / System anatomy

Inside the system

How PGC-DHA handles an admissions enquiry.

Follow the request from the React view through session checks, the Express API, and MongoDB.

Trace the request
Concept renderThe assembly illustrates how a system separates into layers. The walkthrough below follows the actual PGC-DHA implementation.

PGC-DHA / Admissions workflow

Trace an enquiry request.

The React view asks for an enquiry summary. Express validates the session, reads and aggregates MongoDB records, and returns data for the view to display.

  1. React view
  2. Session check
  3. MongoDB query
  4. JSON response
PGC-DHA case study
React client

The enquiry view composes the list and report entry point. It uses permission checks to decide which interface sections to show.

Read source
Application state

The enquiry hook tracks cached results, loading and refresh states, and errors. It requests comprehensive data and date-based breakdowns from the API.

Read source
Express API

The server mounts the enquiry router under /api/enquiries. The comprehensive-data endpoint assembles the data used by the enquiry view.

Read source
Authentication and authorization

The enquiry router runs session validation before handling requests. The shared middleware attaches the authenticated user and session. For example, the student route applies a separate IT role check to registration and progression writes.

Read source Related source
MongoDB data layer

Mongoose defines user records, including student fields and references to class records. The enquiry router uses queries and aggregation pipelines to turn stored records into filtered summaries.

Read source
Reporting

The enquiry report is behind a reporting permission in the client. Its report view fetches filtered records and offers PDF and spreadsheet exports.

Read source Related source