Software Development Articles and Guides | ConcertIDC

ConcertIDC's AIathon: Fostering Innovation and Collaboration in 2024

Written by Sarah Grace Hays | Oct 11, 2024 1:31:50 PM

In our quest to stay at the forefront of industry advancements, ConcertIDC recently hosted a groundbreaking internal event designed to ignite innovation and foster collaboration among our team members. Our goal was to create an environment where creativity thrives, and conventional boundaries are pushed to their limits—all within the exciting framework of an AI-focused hackathon. 
Our event aimed to replicate the success of other hackathons by challenging our team to think beyond the ordinary and develop cutting-edge solutions using AI. This high-energy atmosphere was about solving problems and reimagining what’s possible in our field. 

In addition to driving innovation, we saw the event as a prime opportunity to foster collaboration and networking. In our client-based service environment, employees often miss out on working with colleagues from different projects. We facilitated a melting pot of ideas and perspectives by bringing together individuals with varied skills and experiences. This diverse interaction not only enhances cross-disciplinary learning but also nurtures a culture of teamwork. 

The networking aspect of our event also played a crucial role. It allowed participants to connect with like-minded peers, project leaders, and future collaborators. These new relationships build a company culture that values ongoing collaboration and encourages innovative ventures. 

Our internal AIathon was a key initiative to foster these values and improve our understanding and interaction with a key technology shaping our industry. The AIathon was a grand success, so keep reading to learn more about how our employees took advantage of an incredible opportunity. 

Preparation and Planning

At the heart of our company culture is a commitment to continuous learning and adaptation. As the technology landscape evolves, particularly with the rise of AI, we are dedicated to harnessing these advancements to benefit our clients and internal processes. In 2024, our focus has intensified on effectively understanding and implementing AI tools.

One key component of our educational initiatives is our week-long intensive training sessions held during our CEO’s visits to our offices in India. These sessions enhance our team’s skills and foster a collaborative environment where knowledge is shared and innovation is encouraged.

During her most recent trip in August, we organized a comprehensive program centered around AI applications. Each day hosted workshops, hands-on activities, and expert-led discussions that explored the latest AI technologies and best practices. 

Setting the Stage

The objective for the week was straightforward: to ensure everyone gained a comprehensive understanding of AI. We aimed to explore not just one facet but to provide a holistic view of AI and its applications for both our clients and our internal operations. To achieve this, we organized over ten presentations covering various aspects of AI alongside showcasing our AIathon projects.

Our agenda was packed, and each session was designed to foster engagement and knowledge-sharing. As a result, employees experienced many learning opportunities, gaining valuable insights to empower us to harness AI effectively moving forward.

Pre-Hackathon Activities

We kicked off the promotion of the AIathon in April during our CEO’s spring visit to India, generating excitement and anticipation among our teams. The official announcement was made on June 15, unveiling the guidelines for the competition, which were simple yet crucial:

  • All submissions must adhere to our AI governance policies.
  • Participants could propose ideas for clients they’ve collaborated with or any original concept they have in mind.
  • Teams could consist of up to five members.

Teams formed organically, allowing entrants to collaborate with anyone in the company, unrestricted by department or role. We encouraged senior technical staff to partner with junior team members, fostering mentorship and guidance.

Two teams and two individuals were among the four entrants. We inspired participants to be resourceful and innovative, focusing on creating solutions with potential commercial viability.

The Hackathon Experience

Event Kickoff

The event kicked off on Monday of AIathon week with a brief town hall meeting. Our executive team presented on innovation, motivation, and learning. Following the town hall, the first two teams presented.  
 
Day 2 kicked off with the presentations of the final two entrants. These presentations were especially exciting because the first team consisted of first-year employees, and the final was a long-term employee. Seeing employees engage and participate no matter their experience level is a perfect representation of why this event was important for us.  
Following the final competition presentation, we had an insightful presentation from a new hire who started the week of the AIathon and has worked in AI before. The presentation covered some of the more complex implementations of AI. We ended the day with a presentation on “What’s new in AI.”  
 
At the end of day two, we finished our days with cricket and badminton to close on a positive note focused on bonding and fun. 
 
On day three, our Director of Technology, Kannan K, presented the Microsoft AI toolkit and the first of three presentations about Agentic Frameworks. Day four finished out the presentations about Agentic Frameworks, and day five began with a hands-on session about building your own multi-agent workflow. 
 
We closed the week with a closing meeting, during which we presented the AIathon winners and gave out awards.  

The Hackathon Process

We removed the rules and process for this particular event because we wanted to simulate how the real world works. Technology and AI advance rapidly, and we have to be prepared to adjust to new tools and resources at the same rate.  
 
One team that planned to present had completed their thesis, but when it came time to develop the POC, the idea was mostly obsolete.  We work in this world now, and we are doing our best to prepare and learn accordingly. 

After the announcement of the AIathon on June 15th, teams had until July 15th to submit their concepts for approval. This approval was contingent on protecting customer data and intellectual property. From July 15th til the day of their presentation, teams worked on building out their concept. 

Each person participating took this on in addition to the work they were already doing for clients, which we commend.

Outcomes and Innovations

Final Projects and Presentations

Presentation One: 

  • Team: Karthik Sundaram, Mohamed Salawathulla, and Jegan Murugan -  
  • Project: Skin Disease Analysis System using image recognition from a trained model 
  • Description: Our AI-powered application leverages Azure Custom Vision to analyze and diagnose skin diseases from captured images. The app is designed to assist healthcare professionals and patients by accurately and rapidly analyzing skin conditions. Users can upload pictures of skin ailments, and the AI model, trained on diverse datasets of skin disease images, will provide a detailed diagnosis. This innovative tool enhances early detection and treatment, making dermatological care more accessible and efficient. 

Presentation Two: 

  • Team: Venugopal Chandrasekar  
  • Project: Smart Description builder  
  • Description: This feature provides automatic description suggestions for themes and decks based on user selections. Once users select a theme name and deck name, the system intelligently generates a detailed description, offering relevant insights and content suggestions tailored to the selected items. This streamlines the process for users, enabling them to quickly access pre-written descriptions that can be customized further, ensuring accuracy and coherence with their chosen theme and deck. 

Presentation Three: 

  • Team: Yuvaraja Chinnakaruppan and Joseph Anandaraj 
  • Project: MOP ChatBot - a mobile app for PDF RAG 
  • Description: An AI-powered RAG (Retrieval-Augmented Generation) help chat can significantly enhance the user experience in the health sector by providing instant, relevant answers to common orthopedic questions. The chatbot effectively addresses user inquiries by leveraging a comprehensive knowledge base of orthopedic-related PDF documents. This feature offers 24/7 support, improves patient satisfaction, and gathers valuable insights into patient behavior through the collected data.
Presentation Four: 
  • Team: Vijayakumar Mathiyalagan
  • Project: PDFGpt, Fracture recognition, and Vital engine Chat summarization 
  • Description: PDFGpt is an AI-powered application that allows users to upload PDFs, extract data from these documents, and load the extracted data into ChromaDB, a vector database. Users can then ask questions related to the contents of the PDFs, and the system will provide accurate answers. This application leverages Python (Flask) for the backend, the Llama3.1 model for NLP, ChromaDB for vector storage, and React.js for the frontend.

Judging and Awards

When evaluating each of the projects, we kept the following criteria in mind: 

  • Is there a business use case for the idea? 
  • How quickly could this idea be implemented? 

Keeping those questions in mind, the results for the AI Contest were as follows: 

  • 1st place - Vijayakumar Mathiyalagan 
  • 2nd place -Karthik Sundaram, Mohamed Salawathulla, Jegan Murugan + Venugopal Chandrasekar  
  • 3rd place -  Yuvaraja Chinnakaruppan and Joseph Anandaraj + Chris

While initially intending to reward only the top team or individual, we decided that since all teams presented viable AI solutions, they would all be rewarded. The final order was based solely on the amount of work needed to make the solutions presented commercially viable.  

Impact and Implementation

Two of the four ideas presented have since been presented to the clients they were considered for, and one is in the implementation planning stages with the client. Other project teams are considering how two of these tools could be used on their projects to save their clients time and money.  
 
The idea presented during this AIathon has already begun to impact our clients and internally. The payoff for something created to push our boundaries and help us improve has been impactful. 

Reflections and Future Directions

Lessons Learned

We found that on our first time out for an event like this, there were a few things that worked well, including:  

  • Team members could apply the AI concepts they learned and innovative solutions to real-world challenges to add value to our customers.  
  • Despite the regular project work, the team was able to set aside time to learn AI concepts like LLM and RAG in a short span.  
  • An 8-week timeline to self-learn and develop project ideas enabled the teams to learn and contribute quickly and effectively. 
  • Cross-project collaboration - project team members who were heads down on project work collaborated with like-minded people from other projects to brainstorm new ideas and bond outside their usual work environments.  

Looking Ahead

This AIathon helped strengthen company culture by enhancing teamwork, creativity, and real-world problem-solving skills. It pushed people out of their comfort zones and encouraged experimentation with new technologies in a way that is vital to our work. 

As we move forward as a technology-driven company, we commit to organizing at least one such event every six months. We want to continue to allow opportunities for employees to grow and push themselves by fostering innovation and collaboration through this specific environment. 

In the future, we also want to invite participants from other organizations to allow our team to collaborate with other organizations and exercise how we can explore partnerships to innovate better.

Change is inevitable, and we will continue to strive to step out of our comfort zone.