AI Engineer and Full-Stack Developer with expertise in React, Next.js, Python, FastAPI, Rust, and applied AI systems. Computer Science Engineering graduate from FISAT (2022–2026) with a CGPA of 8.66/10.
Open-source macOS voice dictation tool, an alternative to the $20/month commercial Wispr Flow. Achieves ~390ms end-to-end latency through a progressive-paste architecture: raw Whisper transcription pastes immediately while LLM refinement silently replaces it in the background. Distributed via Homebrew.
Technologies: Rust, Groq API, Whisper, LLaMA, macOS Accessibility APIs, Homebrew
RAG-powered semantic search over PDFs, images, and handwritten notes. Processes 1000+ documents with 95% retrieval accuracy and 80% faster query times than baseline search.
Technologies: Next.js, Python, Docker, Google Cloud, PostgreSQL, Redis, Celery, Gemini Pro
AI-powered data visualization platform that converts raw Excel files into dynamic, embeddable charts in under two minutes. Reduces manual charting time by 90% for non-technical users.
Technologies: FastAPI, React, Recharts, Vite, MongoDB, Pandas, OpenAI API
Full-stack AI hiring platform with semantic resume matching (Pinecone) and automated voice-based interviews (ElevenLabs). Cut recruiter screening time by 46%.
Technologies: FastAPI, Next.js, MongoDB, Pinecone, OpenAI, ElevenLabs, Vector Search
Real-time push messaging platform with message traceability and role-based communication for streamlined academic coordination.
Technologies: React, Flutter, FCM, PostgreSQL
Agentic AI Instagram bot that autonomously curates trending tech news, generates carousel posts using Gemini, and publishes to Instagram via Meta Graph API. Dockerized, runs 24/7 on a 77-minute cadence.
Technologies: Python, Gemini, Meta Graph API, Docker, Pillow, Unsplash API
U-Net model to remove rain from images using MSE+SSIM loss, improving image clarity for vision tasks.
Technologies: TensorFlow, Keras, OpenCV, scikit-learn
ASD Detection from Facial Images using VGG16-based transfer learning, achieving 74.99% accuracy.
Technologies: TensorFlow, Keras, Adam
Sentiment classifier for No-Code vs Code Reddit comments using Naive Bayes and SVM, achieving 80% accuracy.
Technologies: Python, scikit-learn, NLTK
Real-time disaster alert mobile app with geofencing alerts for disasters, reducing emergency response time.
Technologies: Flutter, Firebase, FCM, Google Maps API
An app that allows people to list their leftover food and helps in reducing food wastage by connecting them to nearby NGOs or other people in need. Code available on request.
Technologies: Flutter, FCM, PostgreSQL, Real-time
Dec 2024 – April 2025 - Remote (London, UK)
Engineered a scalable WhatsApp messaging engine using Meta API with AI-driven personalization and webhook-based automation, boosting customer retention by 13%. Deployed a data enrichment engine scraping Google, LinkedIn, and Facebook, adding profile depth by 18%.
Skills: FastAPI, Node.js, WhatsApp API, Web Scraping, MongoDB
Dec 2024 – Jan 2025 - Infopark, Kochi
Enhanced UI performance and responsiveness in production apps. Collaborated on scalable components with senior front-end developers.
Skills: React, Next.js, Tailwind CSS, Figma
IEDC FISAT
Built an agentic job discovery and outreach tool with cold call, cold email, and Instagram integrations under 48 hours.
CSE Department, FISAT
Awarded for highest academic performance in the 2022–26 Computer Science batch.
CSE Department, FISAT
Recognized for the best final-year main project among the 2026 batch.
FISAT
Secured second place in the 2026 batch Mini Project Expo at FISAT.
SAP Labs
Selected through competitive screening for SAP Labs' flagship student scholar program (2026).
Pauly Junior Productions
Secured 3rd place at CineHackAI national film-tech hackathon. Developed Celluloid, an AI-powered integrated film production management system that streamlines end-to-end workflows for filmmakers — from scriptwriting assistance to production scheduling.
FISAT
Secured first place in college-level innovation challenge with a cutting-edge solution that impressed judges and peers alike.
IEDC at UEC College
Emerged victorious in a 24-hour hackathon by delivering a complete full-stack solution under intense time pressure.
IEEE FISAT SB
Dominated the coding arena with clean, efficient algorithms in a high-pressure competitive programming environment.
CS Department, FISAT
Developed an innovative certificate management system that solved a real departmental workflow challenge.
IEEE EdSoc Kerala Chapter
Achieved second place in a prestigious 12-hour hackathon, delivering a functional prototype under tight deadlines.
WIN 6.0, 2025, Kolkata
Successfully presented research on 'AI-Enhanced Form Validator' at a national-level academic conference.
Proceedings of WIN 2025, Springer Lecture Notes in Networks and Systems (LNNS), Vol. 1508
DOI: 10.1007/978-981-96-8998-9_19
Authored book chapter on context-aware AI form validation, presented at WIN 6.0 conference in Kolkata, 2025.
Issued by: Microsoft
Demonstrated foundational knowledge of cloud services and how those services are provided with Microsoft Azure.
Translated ProjectHut documentation into Malayalam to improve accessibility and workflow clarity.
Email: vishnuhari.anup@gmail.com
Location: Kerala, India
GitHub: https://github.com/vishnuhari17
LinkedIn: https://www.linkedin.com/in/vishnuhari17/
Portfolio: https://www.vishnuhari17.dev