Via

ChatGPT, Claude, Perplexity, Via’s upcoming LLM for Cities. LLMs are everywhere, but do you know how they work and how they relate to the rest of AI? We’ve got you covered with the next term in our glossary: 💬 A Large Language Model (LLM) is a highly specialized type of machine learning model. It has been trained on an enormous amount of text (trillions of words from books, articles, code, and the public web), more than the Library of Congress 100,000 times over. Its core function is to predict the next word in a sequence. By doing this at a massive scale, it develops a sophisticated understanding of grammar, facts, reasoning styles, and context, allowing it to generate human-like text, answer questions, summarize documents, and even write code. 💡Analogy: Think of it as "super-mega-autocomplete." Your phone often tries to guess the next word you'll type based on common phrases. An LLM does the same thing, but it's been trained on trillions of words and can predict the next word with an incredible understanding of context. The result is the ability to write a complete, coherent essay, not just finish your sentence. 🔗 Relationship to the other terms: An LLM is a specific product of Machine Learning, which is a part of the broader field of AI. Test your knowledge on the most common lexicon of AI in transit and leave the most interesting use of an LLM you've seen in the comments below! https://hubs.li/Q03Vry8T0

How do you redesign a bus network to be more accessible and equitable — and bring the public along for every decision? 🇳🇿 Hawke's Bay Regional Council did exactly that with 𝐑𝐞𝐦𝐢𝐱. Using live, map-based planning, the team compared route options in minutes, assessed walking access for every neighborhood, and redesigned their network. Now 21% more residents live within 400m of a bus stop. “Without 𝐑𝐞𝐦𝐢𝐱 we would have had to either employ consultants, or use Google My Maps, which would have lacked any kind of rich data. It would have taken ages and we would have had to do a lot of manual work between Excel and Google Maps,” explained Russell Turnbull, Manager of Transport at Hawke's Bay Regional Council. Learn more 👉 https://lnkd.in/ecXg6F_E

To kick off this week, we’re proud to share a new short film capturing the riders, drivers, and community members who make Suffolk County’s paratransit service so special. The cinematography is stunning — but the real beauty is in the stories shared 💖. Watch the full piece here 👉 https://lnkd.in/e-xPqVrs

🚍 Happy National #TransitCareer Day! 🚍 Today, we celebrate the 437,000+ Americans working in public transportation and the millions of jobs it supports in the private sector. At Via, we love the work we do and are so proud to partner with cities and governments around the world to build the future of transit. Check out what it’s like to work at Via’s HQ and a day in the life of Reeya Shah, our VP of Transit Ops! “A career in transit is more rewarding than I could have ever imagined. What we do every day changes people’s lives — that’s what keeps me here.” — Reeya Shah We’re always looking for more transit enthusiasts to join our growing team. If you’re ready to make an impact, check out our career opportunities here 👉 https://lnkd.in/dVrsFp5H American Public Transportation Association #NationalTransitCareerDay #PublicTransit

Today we’re covering a term you might hear thrown around less frequently, but it’s the basis for the type of AI most people interact with most regularly. 🤖 Machine Learning (ML) is the dominant sub-field of AI today. Instead of being explicitly programmed with a set of "if-then" rules, an ML system is "trained" by being shown huge amounts of data. It learns to recognize patterns within that data and then uses those learned patterns to make predictions or decisions about new, previously unseen data. Machine learning already powers demand prediction and route optimization across Via’s deployments! 💡 Analogy: Machine learning is like teaching a child to recognize a bus 🚌. You wouldn’t hand them a manual that says "if it has a metal skeleton, many wheels, and carries people around making regular stops it’s a bus". Instead, when you see a bus you point and say “look there’s a bus!” Over time, when they've seen and named enough buses they can infer something is a bus just by a glance. It’s the same with machine learning systems. You show them hundreds of pictures of different buses that are labeled “bus” (you may have helped with this training by solving CAPCHAs on websites!) and eventually the system builds its own model for identifying patterns. 🔗 Relationship to the other terms: ML is a subfield of AI. It's the engine that powers most modern AI applications, including Large Language Models (LLMs) which we’ll cover next time. Don’t want to wait until next time? https://hubs.li/Q03SPSJt0 💭 For our readers: We used the example of recognizing a bus. What's a real-world pattern you'd love for ML to solve for your agency? (e.g., predicting no-shows, finding new demand hotspots)

We’re kicking off a new series demystifying AI in transit — what it is, how it’s used, and its impact on the sector. We have a lot to say on the topic, but we’re starting with the basics. Over the next few weeks, we’ll tackle the most commonly used terms, explore the rich history of AI in transit, and showcase powerful real-world applications like ridership modeling, AI call agents, safety cameras, and more! There’s much more to come, so stay tuned, but now, without further ado: What is Artificial Intelligence (AI)? AI is the overarching field of science dedicated to making computers and machines perform tasks that normally require human intelligence. This includes things like understanding language, recognizing images, solving problems, and making decisions. Analogy: Think of AI as the entire field of "transit”. Just as "transit" is a huge category that includes specific modes like the bus, tram, car, bicycle, and train, AI is a huge category that includes specific methods and applications. It's a discipline, not a single technique. Relationship to the other terms: AI is the broad category that all the other terms belong to. Want to see the whole glossary now? 👇 https://hubs.li/Q03RJDfh0 We want to hear from you: what's the biggest opportunity or the biggest fear you have about AI in public transit?

🎃 Congrats to the City of Salem, MA — one of Via’s longest-standing city partners — on a major milestone for the Salem Skipper! Since launching in 2020, the Skipper has helped Salem advance its goal of becoming a “car-light” community by residents and visitors with a convenient and affordable way to get around. This month, the service hit its highest weekly ridership ever...no surprise, given Salem’s legendary Halloween season 👻...and recently surpassed 500,000 rides in total. Read more: https://lnkd.in/eCtbPtiU

Here’s to the next generation of leaders reimagining how the world moves. We’re thrilled to welcome 44 MBA students from the UW Foster School of Business to our San Francisco office! As part of their annual Bay Area Trek, the students joined us to explore how Via is transforming transit through modern software and tech-enabled operations. The visit featured product demos, and presentations from our Product & Engineering, Marketing, and Strategy & Operations teams — offering an inside look at how we design, build, and scale equitable transit solutions around the world.

𝐇𝐚𝐩𝐩𝐲 👏 1 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 👏 𝐫𝐢𝐝𝐞𝐬! We’re so proud to power the City of Wilson’s transit service and help riders like Mr. Williams in rural America get where they need to go, every day.

🚗 New in Remix: How bus operator carpools can save you significant costs Operator relief trips don’t carry riders, but they carry costs. Now, Remix Carpools automatically groups relief trips into carpools. That means fewer relief trips. Less waste. And more actual service. Carpooling helps: ✅ Reduce deadhead time ✅ Reclaim valuable driver hours ✅ Save significant operational budget It’s all built into Remix Scheduling. Check it out! https://hubs.li/Q03PVYjJ0 #busoperations #transitscheduling #transitplanning #remixbyvia #reliefvehicles