Breaking Games logoBreaking Games Externship via Extern
Shekhar Sainju portrait

Aspiring Computer Scientist

I am a junior in Computer Science, eager to build a competitive resume through impactful projects and internships.

Work samples

I showcase projects that highlight my skills in AI and document automation, particularly my work with Pfizer's advanced technologies.

  • Pfizer Advanced AI-Powered Document Intelligence Externship
    Pfizer Advanced AI-Powered Document Intelligence Externship
    Pfizer logo

    Pfizer · Oct 2026 · ✅ Verified by Extern

    Pfizer Advanced AI-Powered Document Intelligence Externship

    Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.

    AI & MLPythonDocument IntelligencePresentation Skills
  • SQL & Database Architecture Externship with Breaking Games
    SQL & Database Architecture Externship with Breaking Games
    Breaking Games logo

    Breaking Games · ✅ Verified by Extern · ⏱️ In progress

    SQL & Database Architecture Externship with Breaking Games

    Turn six messy CSVs into an analytics database and a Q4 dashboard that drives real business decisions. SQL + Claude.

    Analytics Database DesignData AnalysisSQL FundamentalsData Storytelling

About me

I am a junior in Computer Science, eager to build a competitive resume through impactful projects and internships.

I am a junior pursuing a Bachelor's Degree in Computer Science. Currently, I am enhancing my skills through an externship with Pfizer, where I focus on AI-powered document intelligence. I aim to build a strong portfolio and improve my competitiveness for future jobs and internships.

Externships

SQL & Database Architecture Externship with Breaking Games

In progress

Experience

Co-Founder

Stealth AI Startup · Nov. 2026 – Present

SQL & Database Architecture Externship

Breaking Games (an Admagic company), via Extern · Oct. 2026 – Present

Deli Associate

Woolworths · Jul. 2023 – Present

Machine Learning Externship

Pfizer, via Extern · May 2026 – Oct. 2026

Apple Foundation Program

Apple, hosted at UTS · Oct. 2026 – Nov. 2026

Education

University of Technology Sydney

Bachelor of Artificial Intelligence (WAM 82.25, Distinction) · Class of 2026

UTS College

Diploma of Information Technology · Class of 2023

Skills

AI DevelopmentDocument AutomationOCRLarge Language ModelsResearch and Analysis

✅ Verified by Extern

Pfizer Advanced AI-Powered Document Intelligence Externship

Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.

AI & MLPythonDocument IntelligencePresentation Skills

Overview

I developed a prototype for AI-powered document intelligence at Pfizer, integrating OCR, LLMs, and RAG to automate PDF workflows. This project deepened my understanding of AI applications in the pharmaceutical sector and allowed me to create a compelling portfolio piece showcasing advanced technological integration.

Pfizer Advanced AI-Powered Document Intelligence Externship

What I've accomplished

I produced runnable Colab notebooks for OCR pipelines and image preprocessing, delivered enhanced scanned-image outputs, and wrote a technical summary explaining tokenization, embeddings, transformer attention, pretraining/fine-tuning, and RLHF.

Project breakdown

View all work

✅ Verified by Extern · ⏱️ In progress

SQL & Database Architecture Externship with Breaking Games

Turn six messy CSVs into an analytics database and a Q4 dashboard that drives real business decisions. SQL + Claude.

Analytics Database DesignData AnalysisSQL FundamentalsData Storytelling

Overview

The externship converted six disparate CSV data sources into a cleaned, modeled analytics database and produced a Q4 dashboard. The work cataloged source schemas, mapped relationships between sales, orders, products, and customers, and defined the transformations required to join the datasets into a single analytics model.

SQL & Database Architecture Externship with Breaking Games

What I've accomplished

I produced exploratory analyses from the assembled SQLite tables, including top-10 lists for units sold and revenue, a count of 99 products, summed Meta ad spend of 37200.94, and 1062 unique checkout sessions.

Project breakdown

On the Breaking Games externship I queried the SQLite database with SQL, producing top-10 lists for units sold and revenue, a catalog count (99 products), total Meta ad spend (37200.94), and unique checkout sessions (1062), plus short reactions.

Google Docs
Access
View all work

DermaScanner

Trained a ResNet18 skin classifier to 94.94% test accuracy and 95.48% macro F1 across 11 classes (7,011 images). Built a YOLOv8n detector (9,026 images) that hands low-confidence cases to ResNet18; shown at UTS AI Showcase 2026.

PyTorchYOLOv8ResNet18Grad-CAMFastAPI

Overview

DermaScanner
View all work

AgentContinuum

Building a Windows tool that auto-resumes Claude Code and Codex sessions after usage limits reset. Wrote the design and threat model across 4 versions (review score 7.5 to 8.5/10), with 81 automated tests passing.

PythonPowerShellC#Windows

Overview

AgentContinuum
View all work

Evident

Built a multimodal CV builder that reads PDF and photo CVs and checks every generated claim against evidence. Designed the architecture and solo-built ingestion, data contracts and the model layer, then led 3 teammates. Used a local Qwen judge model as an independent fact checker alongside Gemini, k

PythonFastAPIGemini VisionQwen (Ollama)
View all work

Intern Insider AU

Built an Australian internship board with community moderation and AI resume tailoring that never invents experience. Fixed a multi-worker rate limiter bypass (N times the limit across N workers) and 3 startup database race conditions. Set up CI running 63 automated tests plus Bandit, Semgrep and pi

FastAPIPostgreSQLRedisClaude APIGitHub Actions
View all work