AI Innovation Bridge Hackathon at Intel Innovation 2023

Ramya_Ravi · ‎10-27-2023

Ramya Ravi, AI Software Marketing Engineer, Intel | LinkedIn

Nikita Sanjay Shiledarbaxi, Product Marketing Engineer, Intel | LinkedIn

Intel Innovation 2023 was a 2-day event held in San Jose, CA, on Sept 19-20. At Innovation, leaders from various organizations talked about how Intel assists developers to make technology more secure and accessible across a wide range of workloads – from client and edge to network and cloud. They also presented breakthroughs in hardware, software, services, and advanced technologies for accelerated development and sharpened competitive edges.

Check out the on-demand content of Intel Innovation 2023.

On both days, Intel hosted an AI Innovation Bridge hackathon that offered an opportunity for AI developers to experience the conference in a unique way.

Learn more about the AI Innovation Bridge Hackathon

It was a “build your own adventure” style hackathon. The participants collaborated with the Intel experts to leverage the labs, sessions, demos, and mentorships at Innovation to get technical insights to optimize their team’s solution. They had free access to the Intel’s latest accelerated hardware architectures, such as:

4th Gen Intel® Xeon® Scalable processors
Intel® Data Center GPU Max Series
Intel® Gaudi® 2 on the Intel® Developer Cloud platform

They also had easy access to Intel’s AI software tools and frameworks, including but not limited to:

The participant projects were categorized into one of the three tracks:

Open Enterprise AI: In this track, participants brought their data/code associated with a solution they were building. The goal was to leverage Intel software and hardware optimizations to enhance the solution’s performance.
LLMs (Large Language Models) in the Fastlane: In this track, participants fine-tuned and deployed a Large Language Model (LLM) on the Intel Gaudi 2 AI accelerator and created a real-world solution using LLMs like LLaMa2 and MPT (Mosaic Pretrained Transformers). They experimented with tasks like Chat, Text Summarization, and Q&A.
Generative AI Startup (Startups Only): This track was tailored exclusively for startups working on Generative AI (GenAI) and LLM solutions where developers could enhance their existing products’ performance using Intel’s latest hardware and AI software optimizations.

The agenda of learning, mentorship, hacking, and socializing featured a pre-innovation Q&A call, a kick-off session, three whiteboarding sessions, and a closing award ceremony. All the members of the winning teams were each awarded the grand prize of an Intel® Core™ i9-13900 Processor.

Introducing the concept of “Bridges”

Bridges were curated mini-adventures designed to help the participants cross the gap from inspiration to implementation, tailored explicitly to various AI workloads and use cases. Each bridge had its own ‘Resource Map’—a curated list of tools, workshops, and expert suggestions that participants could leverage. Along the journey, they could find ‘Expert Guides’—subject matter specialists/mentors willing to offer advice and insights. They also encountered ‘Mini-Challenges’—tasks designed to test their skills and offered chances to earn points or rewards.

During the whiteboarding sessions, participants, along with their mentors, built their own bridge for their team using the below template:

Figure 1: Hackathon Bridge Building Template

A few examples of the bridges are GenAI with LLMs, GenAI Vision Solutions, Traditional ML, Operational AI, AI on the Edge, and Cloud Secure and Ethical AI.

Learn more about the Winners

Following were the winning teams for each of the three tracks:

Figure 2: Winners teams of the hackathon

GenAI Startup – Ocuvera

Ocuvera provides a software-based monitoring system in hospital rooms, so no people are needed to actively watch patients. It sends predictive alerts that let nurses quickly and efficiently respond to patients’ needs before their anticipated movement/fall. At the hackathon, the team aimed to convert 20 billion 3D depth images to infrared images using lower-cost cameras. They were able to accelerate the training phase of the basic YOLOX model using the Intel Gaudi 2 processor.

“Using Intel tools and Intel partnerships, building all of our AI with Intel hardware, we literally are saving patients’ lives doing it side-by-side with Intel.”

Steve Kiene, CEO of Ocuvera

Following the hackathon, we had the opportunity to interact with Steve Kiene, CEO of Ocuvera, and Lucas Sabalka, Chief Data and Analytics Officer of Ocuvera. Watch the interview video.

(view in My Videos)

Open Enterprise AI - Atal Agarwal from Immigration First

Immigration First provides personalized guidance and insights through AI-guided visa assistance for an easy immigration process. It helps you to evaluate your profile, pre-fill your visa application using GenAI, provide recommendations, and guide you through the application process. The hackathon taught the team about various Intel AI technologies and solutions for the LLM software ecosystem.

LLMs in the Fastlane – Csaba Toth from SportsBoard

Csaba was one of the first developers who explored an upcoming low-code LLM tool from Intel to perform language modeling tasks like Q&A and summarization with state-of-the-art models like LLama2.

At the award ceremony of the hackathon, Ralph de Wargny (the director of Intel® Liftoff for AI Startups) made a crucial announcement of Ocuvera becoming the 101st member of the Intel® Liftoff program.

Watch the complete award ceremony video.

(view in My Videos)

All the participants walked out of the hackathon with plenty of Intel gear, the official hackathon T-shirt, and a special '404: Limitations Not Found’ hoodie.

Figure 3: Intel’s Swags for Hackathon participants

What’s next?

We will continue creating collaboration opportunities in fun events like the AI Innovation Bridge Hackathon. If you want access to cloud resources to start building with the latest hardware and software available, you can create an Intel Developer Cloud account. We encourage you to also check out and incorporate Intel’s other AI/ML Framework optimizations and end-to-end portfolio of tools into your AI workflow and learn about the unified, open, standards-based oneAPI programming model that forms the foundation of Intel’s AI Software Portfolio to help you prepare, build, deploy, and scale your AI solutions.