His interest in computers began with programming in Basic in junior high, but the Rice alumnus (B.A. ’01) was equally fascinated by languages and linguistics.
“Rice University was one of the schools where I could study both. I wanted to combine the disciplines, if possible, into machine translation. After completing a lot of Linguistics courses, I realized that universal machine language translation was unlikely due to the resource constraints of the time and inherent differences between languages. It was like having cold water splashed on my face.”
He shifted his focus to programming languages and still recalls the impact of the class he took with Dr. Matthias Felleisen.
“It was the single most influential CS course I took. A lot of the ideas he drilled into us stuck with me through the years. As I’ve matured as a programmer, my personal preference has gravitated toward the two poles of staying close to the metal with C and the more abstract functional paradigm,” he said.
Yantosca interned with a local geophysical company, helping automate some of their processes, then moved into the research lab of Robert “Corky” Cartwright.
He said, “I spent two summers in Dr. Cartwright’s lab. My first summer concentrated on gjdoc, an extension of javadoc for Generic Java. The next summer was on the Dr Java project, an IDE for students. Most of my professional work since has been work-for-hire under NDA, so I’ve been glad to have this for my portfolio.”
Unfortunately, Yantosca graduated into the dot-com bomb of 2001. His portfolio was competing for entry-level jobs against those of out-of-work programmers with a decade of experience, so he fired off an email to a local entertainment studio that was gearing up to launch the Anime Network.
“I asked ADV Films about internships and was hired on. A few months in, one of the directors ran into trouble having to spot-fix all the lip flap matching for a title and asked if I could fix each actor’s part before their respective sessions. That boosted my intern status to a more permanent role.”
He worked in the studio’s DVD production team, assisting in localization and writing the automated dialog recording (ADR) scripts, which involved syncing dialogue with the lip movements on screen and also contextualizing the translation.
“I would get a translation of the Japanese film script, which I would adjust for timing and character voice. For instance, a translator might render a line as ‘How dare you hurt Bob?’ In the context of an action mecha anime, I would tweak the line into something more plausible from a battle-hardened military veteran.
“It was also fun to work on shows where we had to accommodate dialects because they were important to the character’s identity. How would we translate dialect –and the connotations that go with it– into American speech and cultural references? It turns out Houston, Texas, particularly the Pasadena dialect, made a good stand-in for Osaka, Japan.”
When ADV’s forward motion began to slow, Yantosca switched industries. Twice. Turning his writing and technical skills into a purchasing job at a mid-stream oil and gas company, he fell back on his software training. When his boss noticed the automation tools he had developed to make his own tasks easier, Yantosca’s job was expanded to include more programming for the instrumentation and electrical engineering department.
Then his passion for creating new tools led him to BRS Labs, a software development company focused on employing artificial intelligence in the security space.
“A high point of working at BRS was the opportunity to take part in cutting-edge research and development, first as a quality engineer and then on the core development team,” he said. “The most engaging work for me was the audio analysis prototype I got to engineer.
“Without any background in graphics processing unit (GPU) CUDA programming or signal processing, I dove into development with the help of my colleagues, particularly Dr. Kishor Saitwal who lent me his copy of Oppenheim and Schafer’s seminal text. The program extracted certain features from incoming audio data and fed that into the machine learning engine that my colleagues were developing.”
GPUs had been doing the heavy lifting for video games and graphics interfaces for years. Now their power was being tapped for processing data streams. Yantosca wanted to dive more deeply into R&D, but became keenly aware of the gaps in his own knowledge. In 2016, Yantosca traveled to San Jose for NVIDIA’s GPU Technology Conference (GTC).
“I was blown away by what people were doing with GPUs. Dr. Simon Ratcliffe gave one of the most inspiring presentations I saw on his work in South Africa for the Square Kilometer Array to process petabytes of data daily on a shoestring budget.
“It also reminded me of my undergrad passion for natural language processing. I felt with the processing power now at hand in GPUs, we could make some serious headway. But if I wanted to continue in R&D, I would need to expand my knowledge with focused study. That’s when I started considering grad school.”
While exploring graduate programs, he started working in Houston’s entertainment industry again, this time as a Game Server Engineer for Six Foot. Even after enrolling in the M.S. program at the University of Houston, Yantosca continued working for his employer.
“Going back is extremely rewarding and extremely difficult. I could not have gone back without the support of my family. But Ryan Rogers and Dr. Robin Fredericksen at Six Foot were very gracious as well. They allowed me to keep working full time while pursuing grad school part-time: heading first to work, then to UH for classes, and finishing my day in the evening by working remotely via VPN. Most people won’t find employers so willing to work with them around day classes.”
This spring, he doubled his focus on his classes and research into parallel and distributed systems to prepare for his next move, this time into aviation software. He recommends other CS alumni considering a return to the classroom for graduate school really evaluate why they want an advanced degree.
“It isn’t just another piece of paper. Think about the research you want to be doing and what you want to accomplish. Make sure you have your finances in order before you go to graduate school. Even with Houston’s low cost of living, you need to be prepared for the expenses of becoming a student again.”