Nutrigenics.Care

Intelligent clinical nutrition and dietary platform designed to customize wellness insights. Powered by our proprietary Nutrition-GPT model, Nutrigenics translates biometric data and user food logs into real-time diet optimization advice and health analytics.

Nutrition-GPT Clinical Guidelines Diet Coaching Privacy-First
  • Integrates structured medical literature for sound diet suggestions.
  • Tracks macro-nutrients and food composition through conversational chat.
  • Builds personalized nutritional plans mapped to individual health profiles.
  • Maintains offline sandboxing for high privacy and local record preservation.
Grants & Funding

Nutrigenics.Care is supported by over $100K+ in grant funding and is backed by the Microsoft Founders Hub with $150K in technical infrastructure and cloud credits.

Nutrigenics screenshot 1
Nutrigenics screenshot 2
Nutrigenics screenshot 3
Nutrigenics screenshot 4
Nutrigenics screenshot 5
Nutrigenics screenshot 6
AI-powered personalized nutrition and lifestyle platform

VisQ app logo VisQ (Visual Query)

On-device iOS application for composed image and video retrieval using a Qwen3-VL-2B runtime on iPhone. VisQ supports text search, reference-image + edit-prompt retrieval, and explainable "Why This Matched" reasoning for local media search.

iPhone AI Qwen3-VL-2B Image + Video Retrieval Explainability Privacy-First
  • Searches indexed local photos and videos directly on iPhone.
  • Runs composed retrieval with a reference image and natural-language edit instruction.
  • Surfaces reasoning-backed explanation chips for retrieved matches.
  • Adapts recent multimodal retrieval research into a product-facing mobile UX.
App Store Download Now GitHub Source code
Research foundation

Based on our recent work CoVR-R: Reason-Aware Composed Video Retrieval, which brings reasoning into composed retrieval and inspired VisQ's local-first iPhone experience.

Available now on the Apple App Store for iPhone.

arXiv Research GitHub
VisQ screenshot 1
VisQ screenshot 2
VisQ screenshot 3
VisQ screenshot 4
VisQ screenshot 5
VisQ screenshot 6
VisQ screenshot 7
VisQ screenshot 8
VisQ screenshot 9
On-device multimodal retrieval product for iPhone

LawaAgent-Studio

Build & Deploy Your Website’s AI Agent in Minutes. Transform your website into an intelligent, always-available assistant that understands your content deeply.

What It Does

Effectively transforms your website into an intelligent knowledge system.

  • Effortless Setup: Paste one script, live in minutes.
  • Intelligent Crawling: Indexes complex sites into a precision knowledge graph.
  • Smarter Answering: Hybrid Monte-Carlo Search + RAG for accurate, context-aware answers.

Key Features

  • Weekly Data-Driven Insights & SEO Opportunities.
  • Privacy-preserving behavior analytics.
  • Zero maintenance; auto-updates as site evolves.

For Whom?

Businesses, Startups, Creators, Agencies, and Public Services.

No-Code AI Platform

LawaAI-Chat

Trusted AI for UAE-Focused Knowledge Discovery. Accurate answers across legislation, business, education, and public services.

Why LawaAI-Chat?

Solves information fragmentation in the UAE by providing a single intelligent interface for official, verified info.

Differeniators

  • Agentic Intelligence: A system of specialized agents, not just a chatbot.
  • Verified Sources: Grounded in official government portals and trusted datasets.
  • UAE-Specific: Deep understanding of local regulations, multilingual (En/Ar).

Impact

  • 67,000+ queries processed.
  • 2,000+ active users.
  • High accuracy in legislation and business domains.
Conversational AI

Lawa.AI

Agentic AI platform for enterprises and governments. Multilingual, privacy-first AI agents. $130K+ Annual Revenue.

Nutrigenics.Care

AI-powered personalized nutrition platform. Nutrition-GPT engine. Microsoft Founders Hub ($150K).

Real Time Piston Ring Segmentation

Designed and developed in CVPR Lab, IIT Ropar. Real-time detection of rings in pistons for engine assembly quality control. Sponsored by Yamaha Labs, India.

July 2019

Real Time Reinforcement Training of Robot

Used Q-Learning to train a robot to follow a line precisely without explicit programming.

GitHub