Optibus

An optimization problem that took weeks in spreadsheets

Planning a city's public transportation is one of the hardest optimization problems there is. You have to cover thousands of trips a day with a limited fleet, schedule drivers who are subject to labor laws, rest hours and collective agreements, and meet the frequency promised to passengers, and every change at one point ripples through everything else. For years this work was done by hand, in spreadsheets, and a good plan took weeks.

From spreadsheets to a cloud platform: a plan within hours

Optibus, founded in 2014 by Amos Haggiag and Eitan Yanovsky, who met during their studies, built a cloud platform that combines optimization algorithms with machine learning and produces a plan within hours. Just as important, it lets the planner test alternatives (what happens if you add a bus to a line, shift the frequency, or change the shift structure) and see the effect on cost and service before deciding. In June 2026 the company launched Optibus Agent, an AI agent built into the platform that helps with scheduling, driver assignments and real-time operations.

Customers in a thousand cities, and the first public transportation unicorn

The company's customers run public transportation in more than 7,000 cities around the world, according to the company's website (November 2026); in Israel they include Egged, Dan and Metropoline. In March 2021 Optibus raised $107 million, and in May 2022 it closed a $100 million Series D led by new and existing investors including Insight Partners, Bessemer Venture Partners, Verizon Ventures, Pitango and Tencent. The round brought the company's total funding to $260 million and its valuation to $1.3 billion, making it the first unicorn dedicated to public transportation. The round came with the launch of Ridership Insights, a tool that lets transportation planners analyze passenger data directly on a map during network planning. "Optibus is proud to be the first unicorn dedicated to public transportation," CEO and co-founder Amos Haggiag said at the time.

The shift to electric buses as a driver of demand

What increased demand is the shift to electric buses, which adds a new and difficult constraint to the problem: charging times, charger locations and a driving range that varies. The more a fleet is electrified, the less practical manual planning becomes, and this is one of the reasons this field, which looks "dull," has become central.

Optimization, not generative AI

This category shows an application that is not "artificial intelligence" in the popular sense of generating text or images, but optimization: a much older field that deals with finding the best solution within a space of possibilities too large to check in full. Machine learning is added on top to predict actual demand and travel times, rather than relying on a theoretical timetable.

The social benefit, and the question of what exactly to optimize

The social benefit is direct: a more efficient timetable means more trips on the same budget, less waiting time at the stop and fewer empty kilometers. On the other hand, optimization aimed only at cost may hurt lines that serve few passengers but are essential to them. The decision about what exactly to optimize (cost, waiting time or geographic coverage) remains a public policy decision, not a software one.