OpenCode Agent Config
OpenCode agents configured around DeepSeek V4 Flash 0731, GLM-5.3-Flash and GPT-5.6 Luna as backend models; chosen specifically to maximise programming tasks completed per dollar. In progress.
Computer Engineering grad from GEC with hands-on corporate experience in QA Automation. Proficient in Core Java, Python, C++, and JavaScript, with specialized technical expertise in building automation frameworks using Selenium WebDriver and Cucumber (BDD).
I am a Computer Engineer graduated from GEC with a strong foundation in software development, test automation, and full-stack web applications. Driven by curiosity and a passion for open-source technologies, cloud computing, and Artificial Intelligence, I love transforming complex ideas into functional, scalable software.
My technical journey includes hands-on experience in both building and breaking software. During my time as an Engineering Intern at OneShield India, I focused on QA and Test Automation, writing robust test cases for enterprise-level software using Selenium and Cucumber. On the development side, I enjoy building responsive cross-platform applications, having developed a full-stack social networking application using the MERN stack and React Native.
Beyond coding, I thrive in fast-paced, collaborative environments. Whether I’m regularly tackling algorithmic challenges in competitive programming contests or leading a team to manage live events as a Coordinator for our technical festival, TECHNIX, I enjoy problem-solving under pressure and driving projects to success.
If you’re interested in contributing in some open-source projects I work on, feel free to get in touch.
Course and personal work, straight from my GitHub. Featured work will grow here as projects take shape.
OpenCode agents configured around DeepSeek V4 Flash 0731, GLM-5.3-Flash and GPT-5.6 Luna as backend models; chosen specifically to maximise programming tasks completed per dollar. In progress.
Provider-aware prompt-cache optimization and telemetry for OpenCode. Works with DeepSeek V4.1 Flash, GLM-5.3-Flash, GPT-5.6 Luna models.
Automated adversarial prompt generation and synthetic dataset creation. Complements the core FortiPrompt malicious prompt detection component.
A simplified red-team evaluation pipeline for LLMs. Used to test and evaluate prompts and responses within FortiPrompt system.
Our Database Management and Query Processing project in Semester 5. A Reddit clone written in React Native within MERN, prioritizing simplicity and usage ease.
A small snapshot of what I’ve been working on lately, straight from my GitHub profile.
GitHub statistics are temporarily unavailable.
View GitHub profile → (opens in a new tab)GitHub statistics are temporarily unavailable.
View GitHub profile → (opens in a new tab)I write about things I’m learning — programming, design, and whatever else is on my mind.
A short walk-through of the way I approach a new programming language: pick a tiny idea, build it end to end, and let the problems teach the syntax. This is placeholder content — replace it with a real article.
Whether it’s a project idea, a question, or just a hello; my inbox is open. The form below reaches me directly, or find me on any of these platforms.