Software Developer crafting robust digital systems.
Specializing in full-stack engineering with React, Next.js, TypeScript, Node.js, and modern database architectures. Focused on building intuitive user interfaces backed by scalable, high-performance services.
Building software with purpose and technical rigor.
I am a Software Developer driven by solving real-world challenges through resilient software engineering and elegant interfaces.
With extensive hands-on experience in full-stack JavaScript/TypeScript ecosystems—specializing in React, Next.js, Node.js, Express, and database management—I build end-to-end applications that are both reliable under heavy loads and delightful to use.
My background encompasses the complete development lifecycle: from scoping system requirements and architecting database schemas to designing RESTful APIs, securing authentication flows, and deploying to cloud environments.
I take pride in continuous improvement, actively exploring modern architectural patterns, performance benchmarks, and edge-native paradigms to keep engineering standards sharp.
Clean Architecture
Writing modular, self-documenting code with reusable patterns and separation of concerns.
Performance First
Optimizing bundle sizes, database query latency, and rendering lifecycles for instant responses.
User-Centric Design
Bridging the gap between robust engineering and intuitive, accessible user interfaces.
Empirical Software Engineering & AI Assistance
Peer-reviewed research and systematic investigations evaluating how software developers calibrate trust, inspect automated code, and ensure reliability.
Developer Reliance and Verification in LLM-Assisted Software Development: A Systematic Literature Review
Generative AI coding assistants have rapidly penetrated daily software engineering workflows. Yet, software engineering literature frequently examines developer reliance and code verification in isolation. Adhering to PRISMA 2020 guidelines and MMAT appraisal, this review systematically synthesizes 12 empirical studies (2025–2026) alongside four methodological benchmarks.
Crucially, existing empirical findings directly conflict: one cluster of experiments reveals that heavy reliance breeds developer complacency and inflates defect escape rates, whereas another cluster demonstrates that developers review machine output with the same rigorous scrutiny given to human peer code.
Reliance is inconsistently measured in the literature, often reduced to single-item behavioral counts (e.g., autocomplete acceptance rates). Baltes et al. noted this conflation damages construct validity: developers frequently accept suggestions to avoid typing repetitive boilerplate with the intent to refactor immediately.
Sabouri et al. identified a stark intention-behavior gap: developers' stated trust criteria (code readability and perceived correctness) do not align with their real-time evaluation behavior during live IDE programming.
Verification is captured through IDE interaction telemetry (AST diff inspection, test execution, comment-to-code switches) and eye-tracking fixation durations on lines harboring seeded defects.
Tambon et al. categorized 10 recurring LLM bug patterns (hallucinated APIs, execution environment misconceptions). Tang et al. and Khojah et al. showed developers spend more visual dwell time on code explicitly badged as AI-generated.
A sharp empirical split exists across the literature between two distinct experimental findings: Cluster 1 (Perry et al. & Prather et al.) found AI access induced overconfidence and higher defect escape rates; Cluster 2 (Kazemitabaar et al. & Klemmer et al.) found manual problem-solving and self-reported review rigor were preserved.
Differences in task stakes (security vs utility), participant seniority (novices vs industry seniors), and measurement ground truth (seeded test suites vs subjective surveys) account for this divergence.
The review establishes a hierarchy of candidate moderators ranked by empirical validation: 1. Provenance Disclosure (Strongest Support); 2. Developer Seniority (Plausible); 3. Task Stakes (Hypothesized).
Directly manipulating provenance indicators changes how engineers inspect code in IDEs (Tang et al., Khojah et al.). Explicit alerts increase comment-to-code switching and risk mitigation.
Production systems & projects.
Selected applications demonstrating real-world engineering—from real-time WebSocket state synchronization to full-stack microservices and client platforms.
Aurelium
AI-Powered College Learning & Management Platform
Multi-tenant academic platform featuring automated classroom workflows, role-based access, attendance tracking, and an integrated Python microservice for document OCR.
- Express 5 microservices paired with Python FastAPI for document OCR and AI workflows
- PostgreSQL with pgvector for semantic retrieval, Drizzle ORM, and AWS S3/SES storage
Baghchal
Real-Time Multiplayer Strategy Board Game
Digital recreation of Nepal's traditional two-player board game (Tigers & Goats), built with an authoritative server-side rules engine and low-latency WebSocket synchronization.
- Authoritative game engine validating piece placements, leap captures, and dynamic win conditions
- Low-latency Socket.IO rooms with live state broadcast, match replay logging, and Prisma ORM
YourOSHC
Overseas Student Health Cover Comparison Middleware
Insurance middleware and aggregation platform enabling international students in Australia to compare and purchase government-approved health cover across multiple providers.
- Multi-provider policy comparison engine calculating real-time quotes across Australian insurers
- Secure checkout integration with Stripe payment processing and automated policy confirmations
Professional experience.
A timeline of my 2+ years of hands-on software development roles, delivering real value and scalable software.
Full Time Developer
- •Engineering, scaling, and maintaining full-stack web applications and core features across the technical stack.
- •Developing resilient RESTful APIs, optimizing database schemas, and building responsive, high-performance UI components.
- •Collaborating with the engineering team on feature roadmaps, code reviews, and continuous deployment workflows.
MERN Stack Intern
- •Built and tested full-stack features using MongoDB, Express.js, React, and Node.js in an active production environment.
- •Developed responsive user interfaces, integrated frontend state with backend REST endpoints, and resolved edge-case defects.
- •Demonstrated consistent technical growth and problem-solving, earning a direct transition to a full-time developer position in Dec 2024.
Technologies & Tools.
Core technologies and developer tools I use to build fast, scalable, and reliable software.
Next.js & React
App Router, Server Components, SSR, and interactive modern interfaces.
TypeScript
Static typing across components, server handlers, and database contracts.
Node.js & Express
Asynchronous runtime for high-throughput REST APIs and microservices.
PostgreSQL & Prisma
Relational data modeling, schema migrations, and type-safe database queries.
MongoDB & Mongoose
Flexible document data modeling and high-performance aggregation pipelines.
AWS Cloud
Cloud infrastructure, serverless functions, object storage, and scalable hosting.
Socket.io
Low-latency bidirectional WebSocket communication for real-time collaboration.
Docker
Containerized environments ensuring consistency from local development to production.
Tailwind CSS
Responsive layouts, fluid design tokens, dark mode, and modern UI styling.
OpenAI & AI APIs
LLM integration, streaming responses, structured schema outputs, and embeddings.
Zustand & TanStack Query
Client state management paired with automatic background server revalidation.
Git & CI/CD
Version control workflows, automated deployment pipelines, and environment management.
Academic education & achievements.
Formal degree program, university distinctions, and rigorous core computer science curriculum.
Bachelor of Computer Applications (BCA)
Pokhara University, Nepal
Completed a rigorous 4-year undergraduate degree in computer science and software applications at Pokhara University. Maintained academic distinction with a cumulative 3.75 CGPA across the full degree curriculum and recognized on the Dean's List for consistent academic excellence and software engineering performance.
Academic Honors & Distinctions
- Listed in the Dean's List of Pokhara University for academic performance & software engineering excellence
- Maintained a cumulative 3.75 CGPA across the full 4-year undergraduate program
- University Hackathon Winner — Best Full-Stack Web Application (2022)
- Author of Systematic Literature Review on Developer Reliance and Verification in LLM-Assisted Software Development
Core Computer Science Curriculum
Comprehensive foundations covering systems, algorithms, databases, and modern software design:
Let's build something exceptional together.
Whether you have an upcoming project, are looking to hire a full-stack engineer, or simply want to connect—my inbox is always open.