Waypoint Transit Is Letting AI Plan Your City's Next Road

The Macro: City Planning Moves at the Speed of Government

Urban planning in the United States is, to put it charitably, slow. A typical corridor study or transportation planning project takes 18 months to three years from kickoff to final deliverable. The process involves traffic counts, environmental reviews, public comment periods, consultant reports, more consultant reports, and eventually a document that recommends things the city may or may not have the budget to implement.

U.S. municipalities collectively spend around $50 billion annually on planning services. Most of that money goes to engineering and planning consultancies like AECOM, WSP, Kimley-Horn, and Stantec. These firms do competent work, but they operate on billable-hour economics, which means there’s no structural incentive to speed things up. A corridor study that could be done in three months takes a year because the firm has 12 months of revenue to capture.

The result is that American cities are chronically behind on infrastructure planning. By the time a study recommends a new bus route or bike lane, the traffic patterns that justified it may have already changed. Population shifts, new development, changing commute patterns. The planning data goes stale because the planning process takes longer than the conditions it’s measuring.

Technology has been surprisingly absent from this picture. GIS tools like ESRI’s ArcGIS are standard, but they’re data visualization platforms, not planning automation tools. UrbanFootprint and Replica have made progress on data modeling for urban environments. But the actual analytical work of reading crash reports, evaluating sight lines, cross-referencing zoning codes, and writing the planning study? That’s still done by humans with clipboards and Word documents.

The Micro: Stanford Engineers Who Started With a Bus Sign

Waypoint Transit is an AI platform that automates civil infrastructure planning studies. The pitch: what used to take years and cost consulting-firm rates can be done in months at roughly 30% of the typical cost.

Varun Tandon (CEO) and Ryan Johnston (CTO) are both Stanford-educated engineers. Varun has an applied ML background from Microsoft, where he worked on diffusion models and LLMs for Office products including Copilot, Designer, and PowerPoint. He holds both a B.S. and M.S. in Computer Science from Stanford. Ryan’s background is in electrical engineering, also Stanford B.S. and M.S. He previously worked on chip design synthesis CAD at Apple.

The origin story is worth noting. Ryan built novel real-time transit signage for his hometown, which evolved into the broader Waypoint platform. That’s the kind of founder story I find credible. Not “we identified a market opportunity through analysis” but “I was annoyed by a specific problem and started building.”

They came through Y Combinator’s Winter 2025 batch. The team is two people. Their YC partner is Brad Flora.

The product handles several specific planning workflows. Corridor studies use AI-powered geospatial analysis, where planners can run spatial operations using plain language instead of traditional GIS tools. Development reviews automate permit review and report generation against municipal codes. Safety audits identify risk using crash data and infrastructure analysis. The platform connects crash reports, GIS layers, transit feeds, and zoning data into a single analysis environment.

The customer list is already meaningful for a company this young. Santa Fe MPO, Teton County, Walnut Creek, Jackson Hole, Marin County, and Berkshire County are all listed as users. These are real municipalities with real planning needs. Getting government clients at this stage is impressive because government procurement is famously slow and bureaucratic. If they’ve already navigated that process successfully with multiple agencies, the sales motion works.

The competitive picture has two layers. On the consulting side, the big firms (AECOM, WSP, Stantec) are the incumbents. They won’t disappear, but they could see margin compression if cities can do basic planning studies in-house using Waypoint. On the technology side, UrbanFootprint and Replica are the closest comparables, but they’re focused on data modeling rather than automating the planning study itself.

The Verdict

Waypoint Transit is targeting a massive, slow-moving market with a product that appears to deliver real time and cost savings. The founder backgrounds combine AI/ML expertise with genuine civic interest, and the early customer list suggests they’ve cracked the government procurement challenge that kills most GovTech startups before they get traction.

The risk is adoption speed. Government moves slowly, and even a product that saves cities 70% on planning costs has to survive a multi-month procurement cycle. The question is whether Waypoint sells directly to municipalities or partners with existing consultancies who use it to deliver faster results. Both paths work, but they require different go-to-market strategies.

In 30 days, I want to see a case study with actual numbers. How long did a specific study take with Waypoint versus the traditional approach? What did the city save? In 60 days, the geographic expansion question matters. Are they only working with small to mid-size municipalities, or can the platform handle the complexity of a major metro area? In 90 days, watch for whether the big consultancies view Waypoint as a threat or a tool. If Kimley-Horn or WSP starts licensing the platform to speed up their own work, that’s a different business model entirely, and possibly a better one.