Skip to content
Saket Kumar Mall
  • Home
  • About
  • Projects
  • GSoC 2026
  • Contact
Projects

Projects

Engineering work written up as case studies rather than screenshots: what the problem was, what I built, and what it cost.

  • 01

    Adding Spatio-Temporal Dataset Support in Data Catalog

    Merged

    Space-time datasets were invisible in the GRASS GUI. I extended the Data Catalog to read the temporal database directly, list STRDS, STVDS and STR3DS in their mapset trees, and manage them without dropping to the command line.

    • Python
    • wxPython / wxGUI
    • GRASS temporal framework
    • SQL
    • Git
  • 02

    Recoup AI Agent

    Prototype

    An autonomous revenue-recovery agent for failed payments, abandoned checkouts and overdue invoices. It holds one case in five out of its own reach as a control arm, and reports only the incremental difference, with intervals, whether or not that flatters it.

    • TypeScript
    • Next.js
    • Drizzle
    • SQLite
    • Python
    • scikit-learn
    • Vitest
  • 03

    ChronoFlow

    Shipped

    A full-stack productivity platform for meetings, deadlines and leave workflows: a Spring Boot API over a normalized MySQL schema, a hand-built HTML, CSS and JavaScript frontend, and JWT-backed role access.

    • Java
    • Spring Boot
    • JavaScript
    • HTML
    • CSS
    • MySQL
    • REST
    • JWT
  • 04

    Real-Time YouTube Sentiment Pipeline

    Shipped

    A streaming pipeline on Google Cloud that processes multilingual YouTube comments with sub-minute latency and surfaces sentiment trends in an interactive dashboard.

    • Python
    • Google Cloud
    • Pub/Sub
    • Dataflow
    • Streamlit
  • 05

    ATS Resume Ranking Engine

    Shipped

    An embedding-based engine that ranks resumes against a job description, improving match accuracy by about 25 percent over keyword scoring.

    • Python
    • Machine Learning
    • Embeddings
    • SQL
    • Streamlit

Saket Kumar Mall

© 2026 · Built with Hugo