Have you ever wondered what’s beyond the code? ClickIT and Ultrasound AI team of engineers is about to reveal it.
Beyond the Code is ClickIT’s new video series in which teams of developers share what they learned from different projects. For this first episode, the team of Ultrasound AI told us about their incredible work with the healthcare company.
Here’s a bit of what they had to say.
Getting to Know Us: About the Project
Ultrasound AI is a healthcare startup helping detect and treat premature births. As Luis, the team’s project manager, mentioned, “Their goal is to prepare future parents better.” Give them more information regarding the mother’s pregnancy development to identify if their baby is expected to be born prematurely.
The company harnesses the power of artificial intelligence and ultrasound images to achieve its purpose, presenting probable delivery dates for a baby. ClickIT’s part in this noble cause was to develop an application that would present Ultrasound AI’s work to doctors and ultimate users.
“The idea of working in healthcare and using AI through the project really excited us,” stated Luis Alberto, backend developer. Nonetheless, a great sense of responsibility for the project’s meaning and challenges soon joined the excitement.
Challenges and Lessons
Time truly dared this team’s abilities to work efficiently. Like Luis states, “We had a month to build the application and deliver an MVP as it needed FDA’s approval.” This last requirement also meant documenting the project, adding an extra layer to this already-defying mission.
The unusually short development time and the complexity of the application raised the stakes for the members who stood through the test with commitment and strong collaboration.
Karla, leader of the frontend team, shared a different challenge they found in the process. “Another challenging aspect fundamental for the app…was a graph that shows the ultrasounds and the likelihood of a baby being born on specific dates.”
“The graph was crucial for the app… it presented the AI’s results and other required data. So, if the graph was inaccurate, all the work behind it was almost useless,” said Josué, frontend developer. After some consideration, the team opted to go for a model of the graph that would later be modified for a version better suited for the data and user.
The Ultrasound AI team also used Beyond the Code space to discuss other technicalities like the technologies used for development and infrastructure.
“We decided to divide the application. We had the frontend hosted on S3 Bucket, an AWS Service. Most of our infrastructure is on AWS; the rest is on MongoDB Atlas. We used S3 Bucket for frontend, and we employed EC2 instances for the backend,” commented Daniel, DevOps engineer of the project.
“In fact, we also added an autoscaler, as one of the app’s requirements was being highly scalable and available, which are basic characteristics for nowadays apps. We hosted the AI on Lambda and used another Lambda for all the movements required to keep the backend, the Lambda, and the database connected,” he noted.
Beyond the Code: Ultrasound AI’s Success
Looking back on the lessons learned, friendships, and experiences is great joy to the Ultrasound AI team. As Luis remarked, “If I could highlight any characteristic that differentiates Ultrasound AI from other projects is the trust in each team member…Friendship is another relevant aspect of our team.”
“We appreciate each other and also our client. We like him, he’s our friend, and we consider him one of us… I believe those two feelings prevail in our team. Despite all the deadlines and stress, I trust everyone will do their best,” he added.
All the challenges, insights, and responsibilities during the project’s development paved the way for the team’s success. Now, they can chat about it, reflect on what they learned and move on to the next expedition. Let us hope it’s at least half as special and remarkable as Ultrasound AI.
Stay tuned for the next episode of Beyond the Code!