LU students win first place at RowdyHacks with scam prevention technology

Two Lamar University students earned first place in the SWIVEL Challenge at RowdyHacks XII, a 24-hour hackathon held Oct. 3-4 in San Antonio, for developing an artificial intelligence-powered tool designed to detect and interrupt phone scams in real time.

Computer science major Bilal Maqsood and English major Bronwyn Jones developed Canary AI, a system that monitors phone conversations for signs of fraudulent activity and warns users before they fall victim to a scam.

Hosted annually by the Association for Computing Machinery (ACM) at the University of Texas at San Antonio, RowdyHacks brings together students from across Texas to develop technological solutions to real-world challenges. This year's competition drew nearly 400 participants, who worked in teams to design, build and present projects within 24 hours.

For Maqsood, a Spring, Texas, resident who previously participated in RowdyHacks, the competition offered an opportunity to address a problem he frequently encounters.

"I get a ton of scam calls every day, and they are genuinely so annoying," Maqsood said. "I had always wanted to build something to tackle this problem head-on, so when SWIVEL hosted this as a hackathon track, everything just clicked into place."

The team's solution, Canary AI, analyzes phone conversations as they occur, converting speech into text and identifying common scam tactics, including demands for gift card payments, requests for verification codes and threats of legal action. When suspicious activity reaches a designated threshold, the system delivers an audible warning during the call and can send a text alert to a trusted contact.

Jones said the project was designed to address a gap in existing scam prevention technology.

"Most scam protection happens before or after a call," Jones said. "You get a 'Scam Likely' label, or you find out after the money is gone. But the real damage happens in the middle of the conversation, when someone is being pressured to act right now. We wanted to build something that steps in during that moment and gives people a chance to pause."

The pair combined their academic backgrounds to develop the project, with Maqsood leading the technical architecture and programming while Jones focused on researching scam tactics, testing the system, troubleshooting and developing the project's branding and presentation.

"Language is at the heart of every scam, so my English background was a natural fit for studying how scammers pressure people and turning that into scripts we could test," Jones said. "Those tests helped us tune Canary AI to catch the right moments."

The development process presented several technical challenges, including synchronizing live-call data, delivering text alerts and preventing false alarms during legitimate conversations. The team refined the system to evaluate completed sentences rather than incomplete speech and distinguish fraudulent demands from routine communications.

"A tool that flags every call is a tool people turn off, so that balance mattered as much as catching scams," Jones said.

Their work earned first place in the SWIVEL Challenge, which featured 22 competing projects.

"After the second-place team was announced, we thought there was no way we'd take first out of the 21 other projects in our track," Jones said. "When they said our name, I was surprised and honestly couldn't process it at first. We had put everything into those 24 hours, and that moment made all of it feel worth it."

Both students credited their experience at Lamar University with preparing them to collaborate across disciplines.

Jones said her involvement in Lamar's ACM chapter, despite majoring in English, helped her develop an interest in technology and recognize how skills from the humanities could contribute to technical projects.

"The skills I built in my Literature and Writing courses, combined with my time in ACM, gave me the confidence to contribute to something meaningful," Jones said.

Maqsood said working across disciplines strengthened the team's approach to problem-solving.

"Bringing different perspectives together allowed us to look at the problem from both a technical angle and a user experience standpoint," Maqsood said. "Collaborating across disciplines taught us how to communicate ideas faster and build a more well-rounded project."

Following their win, Maqsood and Jones plan to continue developing Canary AI and present the project to SWIVEL representatives at the company's San Antonio office in the coming weeks.

The current prototype protects calls routed through a dedicated phone number. Future development could include integrating the technology directly with phone carriers or mobile applications, expanding its capabilities and making it more accessible to users.

Beyond the competition, both students hope their experience encourages others to pursue opportunities outside their academic disciplines.

"Don't let your major define what you're allowed to explore," Jones said. "Some of the most valuable experiences come from stepping outside your field, whether that's attending a club event, taking on a project, or saying yes to something that feels a little out of reach."

Maqsood also encouraged students interested in technology to participate in Lamar's upcoming Major League Hacking Hacktoberfest Hack Day on Oct. 30. Registration information is available through LU Hub.

"You don't need to know everything before you start," Maqsood said. "Hackathons are all about learning on your feet, having fun, and building something cool with your peers. The worst that can happen is you learn a ton along the way."