Anushiv Prakash@prakashanushiv36106
1
Events attended
1
Submissions made
India
1 year of experience
About me
Hi, I'm Anushiv Prakash, an aspiring software engineer from India with a strong interest in artificial intelligence, full-stack development, and building products that solve real-world problems. I enjoy taking ideas from concept to deployment, whether it's developing AI-powered web applications, experimenting with large language models and AI agents, or creating tools that make information more accessible and useful. My recent work includes projects focused on AI-assisted decision making, modern web technologies, and user-centric application design. Beyond coding, I've been actively involved in debating, event management, mass communication, and editorial work during school, experiences that strengthened my communication, leadership, and teamwork skills. I believe great technology is built through collaboration just as much as technical expertise. I'm passionate about open source, hackathons, and continuously learning emerging technologies. I enjoy exploring new frameworks, contributing to developer communities, and challenging myself with projects that push me beyond my comfort zone. As I begin my journey in computer science, my goal is simple: keep learning, keep building, and create technology that has a meaningful impact. I'm always excited to connect with fellow developers, mentors, and innovators who share a passion for AI and software engineering.
š¤ Top Collaborators
š¤ Latest Submissions
DescribeX
DescribeX is an AI-powered video captioning platform built for creators, educators, developers, and accessibility advocates. Users upload a short video (30 seconds to 2 minutes), and the system automatically analyzes the visual content before generating four distinct caption styles: Formal, Sarcastic, Humorous (Tech), and Humorous (Non-Tech). The application follows a two-stage AI pipeline. First, a multimodal vision model extracts an accurate scene understanding from sampled video frames. That structured description is then passed to a dedicated language model that produces high-quality captions in multiple writing styles. This separation improves consistency, reduces hallucinations, and makes the system easier to extend with additional styles in the future. DescribeX includes an integrated caption editor, TXT export, and automatic burned-caption video generation through an FFmpeg-based rendering pipeline. The frontend is built with Next.js, while the backend uses FastAPI with a modular provider architecture that supports multimodal inference through Fireworks AI. The engine performs video validation, intelligent frame sampling, prompt construction, caption generation, formatting, and export in a complete end-to-end workflow. The project was designed with accessibility, simplicity, and extensibility in mind. Rather than generating only generic subtitles, DescribeX allows creators to instantly adapt the tone of their content for different audiences while maintaining a streamlined user experience suitable for both technical and non-technical users.
13 Jul 2026
