Browse applications built on modern technologies. Explore PoC and MVP applications created by our community and discover innovative use cases for modern technologies.
Pre-publish multimodal video optimization platform. Uses AMD Instinct accelerators, MiniCPM-V, and DeepSeek to audit TikTok/Reels UI safe-zones, fix pacing, and rewrite narrative hooks before creators publish, preventing wasted algorithmic reach.
Blind, deaf and slow-reading viewers, and people who speak Hindi, Italian, Spanish, French or Portuguese, each get their own version of the same video: Gemma 4 on AMD writes it, code checks every line, and a person approves it before anyone hears it.
Film a manual operation with a phone. Gilbreth drafts the two-hand (bimanual) chart with Gemma 4 on AMD GPUs, flags method problems and estimates the seconds per unit you can save. A methods engineer reviews and approves it.
It is a boardgame wiki on steroids that can help you play through intuitive videos, and chatbots.
One workspace that takes a fashion product from a sketch to a 3D prototype, campaign content and a market-tested production decision, all guided by your brand's DNA.
Crash Guard uses a fine-tuned YOLOv8 AI model to detect traffic accidents from camera feeds, record video evidence, and send real-time alerts with GPS locations to emergency responders through a connected Flutter mobile and web application.
Upload a football match video. Trivela tracks every player and the ball with Computer Vision and turns it into 25+ stats. With an AI chatbot powered by AMD using answers your questions about the match. No trackers, sensors or extra tools, just one video.
SnapSell turns narrated inventory videos into reviewed resale listings, researches evidence-backed asking prices, prepares marketplace listing kits, and helps sellers manage buyer replies, pickups and recorded sales with AMD-powered AI.
Spectra is a real-time music visualiser you drive by voice. AssemblyAI Universal-3.5 Pro turns plain speech into 238 app commands — play a track, switch visual style, tune audio — code-switching across 18 languages and returning each speaker's own script.
Cutback is a voice-controlled editor for short video. Upload a clip, get a transcript, then say what to cut, caption, or reframe. The preview and the exported file are the same edit.
VocalScout is an AI-powered video assistant using RAG vector search, AWS Serverless, and WebSockets to answer queries with timestamp-accurate video segments, real-time voice synthesis, and multilingual chat.
DRIFTZERO autonomously analyzes, plans, migrates, validates, and safely delivers SDK and API upgrades from one version to another, using IBM Bob for intelligent migration reasoning.
A2Z Nexus is an AI-powered multi-agent workspace for developers and engineers to ask technical questions, analyze projects, test code, generate documentation, identify improvements, manage deadlines, and assess project readiness from one unified platform.
MediFact listens to live disputes, identifies each speaker, extracts their claims, and audits them against your evidence — producing a report marking every claim supported, contradicted, or unverified, with verified quotes.
A camera on a SiMa.ai Modalix chip that remembers everything in a room and answers out loud: "Where's my red mug?" "Who took the laptop?" Segmentation, pose, tracking and a vision-language model on one sub-10 W chip, with no cloud at all.
A real-time edge AI fashion stylist on Modalix that detects and tracks and selects people, analyzes visible outfits, understands occasion and cultural context, and delivers personalized conversational recommendations.
STICK is a conversational AI that turns spoken ideas into professional, context-aware, editable presentations—without lengthy prompts. Just speak, refine your story, and STICK builds the slides.
Prototyped visual kit inspection with physical actuation. Snapdragon X Elite (on-device VLM) + Genie X ollama gwen2.5vl:7b + Arduino UNO Q (solenoid latch). Fails closed by design.
ZORN SIGHT is a fast offline-first multimodal AI system for Intel edge devices, combining chat, image and video understanding with smart preprocessing and optimized local inference.
Voice-controlled dual-arm manipulation in MuJoCo on Intel Core Ultra 7 358H. PoseNet perception is benchmarked with OpenVINO on CPU, iGPU and NPU, but the demo runs oracle mode. Four scripted skills, 20-seed evaluation. 59 ADRs.
An SO-101 arm inspects LEGO bricks, classifies each as defective or good with Anomalib on an Intel NPU, and sorts it into the matching plate. The verdict reaches the robot as an English instruction that conditions a fine-tuned SmolVLA policy.
A real-time edge AI pipeline that decodes RTSP video, runs YOLOv8 object detection on SiMa Modalix hardware, and streams synchronized video and detection overlays to Neat Insight.
A bimanual robot arm on Intel hardware, rigorously tested for whether voice control actually works — honest results included, even the failures.
An AMD-powered edge AI security agent that detects suspicious events, understands activity over time, assesses risk, and generates real-time actionable incident alerts.
Ohun is a real-time communication app that translates voice calls, chats, and voice notes so people can communicate naturally across languages.
Judgment is the bottleneck, not editing. Brandy, a voice agent on AssemblyAI, walks a B2B founder from expertise to a script that passes 14 rules and a finished video, and in Agentic Mode she does it with you, confirming every paid action out loud.
AI-powered YouTube content factory that automates script generation, narration, visual assets, video assembly, and publishing through a single end-to-end pipeline.
MeetPlay is a functional, deployed video-meeting app that turns the live transcript into the game layer and the recap.
Askable turns any video into a source you can question in plain language. Paste a YouTube link or upload a file, ask anything, and every answer links to the exact second it came from.
SovereignSwarm is an autonomous video captioning engine using Fireworks AI and OpenCV to extract chronological frames and generate structured, multi-tone captions (Formal, Sarcastic, Humorous Tech, Everyday) in an enterprise-ready JSON format.