Final-year Computer Science student at BITS Pilani Dubai, building at the intersection of AI, product, and operations.

• Head of Operations at R13 Labs, an early-stage EdTech startup
• Experience building AI-powered products end-to-end — from UI design to ML pipelines to deployment
• Strong focus on RAG systems, applied ML, and full-stack product development
• Interested in AI Systems, Agentic Workflows, and Operations-driven Product Building

AI academic study hub featuring an integrated AI assistant (Ada), smart schedule management, grade tracking, Google Calendar sync, and study planning — currently in private beta with 500+ users ahead of iOS and Android launch.

• Built the PWA using React 18, TypeScript, and Vite with Tailwind CSS and shadcn/ui components, backed by Supabase (PostgreSQL + Deno Edge Functions) and deployed at a custom domain
• Co-designed the entire mobile app UI in Figma and converted all screens into a Flutter/Capacitor codebase using AI-assisted tools
• Integrated AI assistant "Ada" with OCR-based schedule parsing, contextual study insights, and Google Calendar OAuth sync with persistent token management
• Recognised as Innovative Education Platform of the Year — CorporateLiveWire Innovation & Excellence Awards 2026 (UK)

Dual-dataset machine learning pipeline detecting student stress from wrist-worn sensor signals.

• Built using WESAD and DEAP physiological datasets, achieving 91.33% accuracy
• Engineered 10 physiological features from EDA, BVP, and skin temperature signals with 5-fold stratified cross-validation
• Developed a cross-dataset bridge connecting stress detection to emotion zones, generating real-time personalised binaural beat and haptic intervention prescriptions via a live Streamlit application

Token-level uncertainty quantification pipeline for detecting hallucinations in RAG systems.

• Designed using Information Gain, KL Divergence, and Confidence Drop metrics computed over GPT-2 probability distributions
• Built a logistic regression composite model achieving 0.734 AUROC on RAGTruth
• Evaluated across 6 LLMs including GPT-4, GPT-3.5, Llama-2 variants, and Mistral-7B, with zero-shot cross-domain transfer to HaluEval yielding 0.925 AUROC

Production-grade web application for separating audio tracks into 6 isolated stems (vocals, drums, bass, guitar, piano, other) using Meta's HTDemucs model.

• Engineered an audio separation pipeline using HTDemucs achieving high Signal-to-Distortion Ratio (SDR), with FFmpeg preprocessing to normalise and convert arbitrary audio formats into standardised model inputs
• Architected an asynchronous FastAPI REST API handling file uploads, background job processing, and real-time progress polling — returning a job ID immediately so the frontend can track long-running inference tasks
• Implemented dynamic hardware acceleration detection across NVIDIA CUDA, Apple MPS, and CPU backends, making the application hardware-agnostic out of the box

Head of Operations — R13 Labs India Pvt Ltd (Sept 2025 – Present)

• First operations hire at a 12-person EdTech startup; designed operational infrastructure spanning communications, project management, CRM, finance visibility, and HR onboarding
• Co-developed a 26-week company-wide Gantt roadmap spanning Technical, Marketing, Operations, and Finance workstreams tracking app launch milestones, university outreach, and a 10,000-user growth target across 4 phases

Software Engineer Intern — TechnoCit Software Solutions (Jun 2025 – Oct 2025)

• Developed an agentic banking chatbot and multi-country web scraping systems using n8n
• Engineered RAG pipelines using vector embeddings and semantic similarity search

• MIF Finalist — Mookerji Innovation Fund · AWS & GCP Cloud Credits Recipient
• CorporateLiveWire Innovation & Excellence Award 2026 (UK)

LinkedIn: zayaanalimk · Email: zayalimk@gmail.com · Dubai, UAE