अथ / A BEGINNING
Vishal Mahajan portrait
The guy behind the screen01 / 02

Hello, I am

Vishal Mahajan.

A collection of things I've made, things I've figured out, and things I'm still trying to make sense of. The lore is ongoing.

Find me elsewhere

कर्म / Work & ExperienceView all

Experience & Impact

My professional journey, roles, and contributions.

Backend Developer Intern

The Good Game Theory

Employment
Internship
Dates
May 2025 – July 2025
Location
Remote

Technologies

Node.js · Express.js · MySQL · JavaScript · EJS · Git

Contributions

  • Engineered and deployed RESTful API endpoints with integrated role-based access control (RBAC) and CSRF protection, enabling secure delivery of dynamic content within an MVC architecture.
  • Migrated 17+ interactive EJS dashboards from a legacy PHP backend to Node.js, improving maintainability and establishing a modular, reusable component structure.
  • Optimized backend code by applying DRY principles and collaborating with cross-functional teams through Git-based workflows, reducing integration defects and improving deployment efficiency.
सृजन / ProjectsView all projects

Projects, Research & Case Competitions

A collection of things I've built, problems I've explored, and challenges I've competed in.

Industry ProjectsJul'25 - Jun'26

H-adGen: Framework for Hyperlocalized Personalization of Video Advertisements

A multimodal AI framework that localizes pre-recorded video advertisements by replacing spoken brand and location-specific entities while preserving the original speaker’s voice and visual identity. Integrates speech recognition, zero-shot voice cloning, lip synchronization, facial restoration, and multilingual content moderation to generate regionally tailored advertisements in Hindi, Marathi, Tamil, and Telugu.

Cere Labs logoCere Labs
Research Paper

H-adGen: Framework for Hyperlocalized Personalization of Video Advertisements

A multimodal AI framework that localizes pre-recorded video advertisements by replacing spoken brand and location-specific entities while preserving the original speaker’s voice and visual identity. Integrates speech recognition, zero-shot voice cloning, lip synchronization, facial restoration, and multilingual content moderation to generate regionally tailored advertisements in Hindi, Marathi, Tamil, and Telugu.

Contributions

  • Addressed the scalability and localization challenges of celebrity-led advertising by developing a framework that adapts a single brand campaign for different retailers, locations, and promotional offers.
  • Designed a brand-to-retailer personalization workflow enabling MSMEs to generate localized advertisement variants without requiring a separate video shoot for every variation.
  • Architected a two-stage multimodal AI pipeline integrating 5+ components across speech recognition, voice generation, lip synchronization, facial restoration, and content moderation.
  • Developed spoken-entity personalization capabilities to replace brand names, location references, and promotional offers while preserving the original speaker’s voice identity and synchronized lip movements.
  • Enabled multilingual advertisement localization with support for 22 Indic languages, targeting regionally tailored advertising across diverse linguistic markets.
  • Engineered an evidence-based benchmarking framework using 640 speakers across two language families to evaluate model suitability for multilingual speech applications.

Recognition

Colloquium 2026
Second Prize
Recognized with second prize at Colloquium 2026.

Gallery

Second Prize · Colloquium 2026

Second Prize · Colloquium 2026

Colloquium 2026 award recipients with their certificate

Recognition

Project Completion Certificate · Cere Labs · June 2026

Project Completion Certificate · Cere Labs · June 2026

Cere Labs project completion certificate for the H-adGen project
Case Competitions2026

Castrol PowerUp 4.0 - Supply Chain Track

Developed a data-driven supply chain planning methodology for a multi-stage network connecting plants, mother hubs, Clearing and Forwarding Agents (CFAs), and customers. Combined inventory planning with Mixed-Integer Linear Programming (MILP) to optimize production and distribution decisions, evaluate capacity constraints, and analyze cost and demand-fulfillment trade-offs, earning National Runner-Up recognition in the case competition.

Castrol logoCastrol
LinkedIn Announcement

Castrol PowerUp 4.0 - Supply Chain Track

Developed a data-driven supply chain planning methodology for a multi-stage network connecting plants, mother hubs, Clearing and Forwarding Agents (CFAs), and customers. Combined inventory planning with Mixed-Integer Linear Programming (MILP) to optimize production and distribution decisions, evaluate capacity constraints, and analyze cost and demand-fulfillment trade-offs.

Key Highlights

  • Supply Chain Network Modeling: Structured production, inventory, and distribution decisions across plants, mother hubs, CFAs, and customer demand.
  • Inventory Planning: Developed an inventory norm methodology using average demand, forecast variability, lead times, safety stock, reorder points, and days of cover.
  • MILP Formulation: Formulated a mathematical optimization model to minimize modeled production, transportation, inventory, and unmet-demand penalty costs while satisfying operational constraints.
  • Production & Capacity Planning: Incorporated fixed 25 kL production batch sizes and plant- and category-level capacity constraints into the planning model.
  • Scenario Analysis: Evaluated a 70% Overall Equipment Effectiveness (OEE) scenario to assess its impact on effective production capacity, modeled costs, and unmet demand.
  • Bottleneck Analysis: Interpreted optimization outputs to identify structural capacity bottlenecks and evaluate production planning and resource allocation decisions.

Optimization Tool Development

  • Excel-Driven Workflow: Implemented spreadsheet data import to populate the optimization model.
  • Configurable Constraint Engine: Enabled users to add custom constraints and modify or remove default constraints for flexible scenario evaluation.
  • Optimization Backend: Developed the backend using Python and FastAPI, with PuLP for MILP formulation and the CBC solver for optimization.

Technology Stack

Python · Pandas · PuLP · CBC Solver · FastAPI · Excel · Mixed-Integer Linear Programming

Recognition

Competition result
National Runner-Up
Castrol PowerUp 4.0 Supply Chain Track Case Competition.

Gallery

National Runner-Up confirmation · Castrol PowerUp 4.0

National Runner-Up confirmation · Castrol PowerUp 4.0

Email confirming the National Runner-Up result in Castrol PowerUp 4.0
चिन्तन / BlogsView all posts

Thoughts, experiments & things I learned

Writing about technology, projects, and the occasional rabbit hole.

Technology4 min read

A Developer CLI Portfolio That Lives Inside Your Terminal

How I built a portfolio that runs entirely in the terminal using npx. Features, design, and the tech stack behind it.

Read article
विद्या / Education

Education & Learning

My academic background, qualifications, and continued learning.

मस्ती / Side Quests

Outside the syllabus.

Because there's more to life than deadlines and deliverables.