What we've built.

A drone detector in field trials, AI that protects children online, and video tools that run on small devices instead of server farms. Each page explains the problem, what we built and what it achieved, with the evidence linked.

DroSonic test team at a table of equipment beside a gravel road on a test range, with service members watching the sky

Detecting drones by the sound they make

Small drones that fly on fiber-optic cable or on autopilot give off no radio signal, so most detectors miss them. DroSonic listens for their propellers instead, on low-cost sensors that send nothing and need no cloud.

Read about DroSonic

Diagram of EdgeVidSum. A video service provider sends thumbnail containers and video segments over the network to a Jetson Nano edge device, where a lightweight 2D CNN uses user preferences to generate timestamps for a personalized fast-forward summary player

Personal video highlights, made on the viewer's own device

A viewer picks what they care about and gets a fast-forward summary of a long video, built on a small device. The video and the viewer's preferences never go to a server.

Read about EdgeVidSum

EdgeAIGuard architecture. An input layer of digital communications, instant messaging and social media platforms feeds an edge processing unit with data ingestion, preprocessing and a message-passing network; Intervention, Context and Sentinel agents use an LLM engine with embedding generation and inference, local history cache and pattern memory, and a protection layer with a threat handler, intervention module and alert system

Catching online threats to children before they escalate

A team of AI agents that reads messages close to the user, spots grooming and exploitation aimed at minors, and steps in within a tenth of a second, without sending private conversations to the cloud.

Read about EdgeAIGuard

LTC-SUM architecture. An HLS IIS server holds M3U8 playlists, segment representations and thumbnail containers; the HLS client's thumbnail containers analyzer finds preferred events, requests a summary, and a segments summarizer module assembles the video summary for the HLS player

A personal highlight reel for every viewer, without a server farm

Each viewer's device finds the moments they care about in a full-length film or match and builds their own summary, so a service can offer personal highlights to many viewers at once without processing video for each of them.

Read about LTC-SUM

FRC-GIF process. Input videos such as football, basketball and tennis; user selection of movie and scene; download LTC; extract and rank thumbnails; recognize events; key thumbnails ordering producing personalized thumbnails; then download segments and generate animated GIFs

Thumbnails and GIF previews that match each viewer's taste

Previews sell videos. This system picks the moments a particular viewer likes in a full-length sports video and turns them into a thumbnail and a short animated preview, using a fraction of the computing of earlier methods.

Read about FRC-GIF

System architecture. An HLS server holds video segments and audio files in several formats; the HLS client's music genre classification module and animated GIF generation module request audio, find the preferred segment and return animated GIFs to the user interface

Finding the best moment in a video by listening to it

Sound often tells you where the big moment is faster than pictures do. This system reads a music video's audio track, finds its climax and turns just that moment into a preview.

Read about Audio-driven highlights

Architecture with an HLS server side holding M3U8 playlists, segment representations and thumbnail containers, connected over the network to an HLS client side with a persistent broadcast connection, thumbnail containers analyzer and HLS player that requests the trailer segments

Personal movie trailers, without collecting personal data

Every viewer sees the same trailer today. This system builds a trailer around the action each viewer prefers, on their own device, so their tastes never have to be shared.

Read about Personalized trailers

Monitor showing a chat web interface and a terminal with jtop reporting GPU, CPU and memory use on a Jetson Orin Nano Developer Kit

A private AI assistant that never leaves your building

Modern AI assistants running on a small computer on your own desk, with a familiar chat window, so documents and questions stay in-house.

Read about Private on-device AI assistants

Have a perception problem that has to run on real hardware?

Tell us what you need to detect, where it has to run and what it runs on. You'll get a reply from the founder.

Email gmujtaba@koraiio.com

Email: gmujtaba@koraiio.com
Founder: Dr. Ghulam Mujtaba
Product: DroSonic
Based in: Denver, Colorado