Nitesh Soni

Software Engineer specializing in AI/LLM, Backend Development, and Secure Systems
Sydney, AU.

About

Highly motivated Software Engineer Intern with a strong foundation in backend development, AI/LLM, and secure system architecture. Proven ability to deliver high-impact solutions, optimize performance by up to 35%, and enhance system reliability for over 100,000 requests and 1,000+ users. Eager to leverage expertise in GraphQL APIs, microservices, and robust security protocols to contribute to innovative software engineering challenges.

Work

Vyaasa.ai
|

Software Engineer Intern

Remote, Australia

Summary

Currently developing and deploying production-grade GraphQL APIs and scalable backend systems to manage authentication, analytics, and document processing for internal users.

Highlights

Shipped production GraphQL APIs using NestJS and PostgreSQL, powering authentication, analytics, and document management for 100+ internal users.

Engineered and implemented Redis and BullMQ job queues, processing over 100,000 asynchronous requests and significantly improving system reliability and throughput.

Deployed secure file pipelines leveraging AWS S3 and pre-signed URLs for PDF validation and parsing, resulting in a 30% reduction in upload errors.

Developed a graph-based code query system (MCP server) for internal dev tooling, enhancing codebase navigation and reducing LLM token usage by 30%.

Containerized critical services using Docker, MinIO, and PostgreSQL, establishing environment parity across development and production environments.

Australian Learning Institute
|

Software Engineer Intern

Sydney, NSW, Australia

Summary

Enhanced the Learning Management System (LMS) through SCORM module integration, robust security implementation, and critical performance optimization.

Highlights

Enhanced the LMS by integrating SCORM modules, thereby unlocking access to over 200 new course modules for students.

Designed and implemented JWT-based access control, effectively securing over 1,000 student accounts.

Optimized critical SQL queries, resulting in a 35% reduction in page load times on high-traffic endpoints.

Education

University of New South Wales
Sydney, NSW, Australia

Bachelor of Engineering

Computer Science

Skills

Languages

Python, Java, TypeScript, JavaScript, SQL.

AI/LLM

LangChain, Langfuse, RAG, Prompt Engineering.

Frameworks/Tools

NestJS, FastAPI, Next.js, React, Express, GraphQL, REST, TypeORM.

DevOps/Infra

Docker, GitHub Actions, Azure, AWS S3, MinIO, PostgreSQL.

Auth/Security

JWT, Firebase Auth, Google OAuth.

Others

Git, VS Code, Postman, Octokit, Linux, Agile.

Projects

Invoice Management System

Summary

Architected and developed a microservice for invoice compliance validation against Australian tax law, featuring robust security and a modular design, which was recognized as a top project at Macquarie Uni.

Bug - PR AI Assistant

Summary

Developed a two-agent orchestration system for generating pull request plans and identifying relevant code files using LangChain with various LLMs, enhancing developer productivity.

HealthFactAI – UNESCO Hackathon

Summary

Built a real-time health claim fact-checking platform for the UNESCO Hackathon, leveraging allow-listed sources and advanced NLP techniques to ensure data consistency.