How a Firmware Engineer Ended Up Inside a Sewer
Every city in America has the same problem and almost nobody talks about it. The pipes installed after World War II - the ones carrying sewage beneath every street - are hitting the end of their designed lifespan at the same time. Rosenthal's phrase for this: "More frequent failures and increased costs." His solution: stop pretending a human with a clipboard is the best approach.
Rosenthal came to sewers the way most good founders arrive at their best ideas - sideways. After a BS and MS in Electrical and Computer Engineering from Carnegie Mellon, he went to Fitbit as a senior firmware engineer, helping lead the team that built one of the first Android tablets. Then Touch Revolution, working on industrial and consumer touchscreens. Then Centosette, a data analysis company built around municipal water systems, which he co-founded and where he served as CTO before its acquisition by TPK - the world's largest touch panel manufacturer - in 2010.
Along the way, something stuck: municipal water infrastructure is complicated, underfunded, and almost entirely managed by legacy tools. Around 2019, while taking online courses in machine learning as a side project, Rosenthal and his co-founder Billy Gilmartin started asking a specific question - what if you pointed computer vision at sewer inspection footage? The industry was spending billions on inspections, generating massive video archives, and doing almost nothing with them computationally.
"Legacy incumbents offer on-premise or on-truck software that has seen very little innovation in the last 20 years."- Matthew Rosenthal, SewerAI
SewerAI's early sales pitch was clever in its specificity: they offered potential customers a free review of their existing inspection video archives. The AI found defects the inspectors had missed. The demos sold themselves.
That approach helped SewerAI land clients that would impress any enterprise SaaS founder - the City of Houston, San Francisco, Toronto, the Knoxville Utilities Board. Not startups or government pilots. Real infrastructure customers with real procurement cycles, real compliance requirements, and real operational stakes.
Pioneer, AutoCode, and the Digital Twin Under Your Feet
SewerAI's platform has two main faces. Pioneer is the end-to-end inspection management layer - field inspectors upload footage to the cloud, annotate issues, and project managers use that data to plan repairs. AutoCode is the AI engine running underneath: it automatically tags defects in pipe and manhole footage, generates 3D models from standard camera footage, and dramatically compresses the time and expertise required to process inspection data.
The 2024 launch of Smart Project Builder extended SewerAI's reach further into the infrastructure lifecycle - converting inspection and GIS data into risk-prioritized rehabilitation project plans. Rosenthal's framing for why this matters is precise: "Utilities are sitting on mountains of inspection data, but that data alone doesn't fix pipes."
"Our products streamline field inspections and data management, enabling clients to proactively manage infrastructure instead of reacting to emergencies."- Matthew Rosenthal, SewerAI
The platform is compliant with NASSCO PACP standards - the industry's inspection coding protocol - and SOC 2 Type II certified. For municipal procurement teams, those aren't nice-to-haves. They're table stakes, and SewerAI cleared them early.
$50 Billion Problem, $18.5 Million in Backing
Rosenthal tends to frame SewerAI's market not in SaaS terms but in infrastructure ones. The number he returns to: U.S. municipalities spend approximately $50 billion each year just maintaining aging pipe infrastructure. That's before any capital projects. That's just maintenance. And roughly 300 million of the one million miles of U.S. sewers need annual inspection and cleaning.
Series B - June 2024
$15M led by Innovius Capital, with participation from Emerald VC, Epic Ventures, Suffolk Technologies, Bentley Systems, and Burnt Island Ventures.
Individual investors include Zachary Bookman, founder and CEO of OpenGov, and CEOs from several AI and computer vision companies - a roster that reflects how seriously civic tech insiders are watching SewerAI's trajectory.
Total capital raised: $18.5M. Revenue more than doubled year-over-year entering the round.
"U.S. municipalities spend about $50 billion annually just to maintain their aging pipe infrastructure. Our software has saved tens of thousands of engineering hours and tens of millions of dollars for our customers."- Matthew Rosenthal, SewerAI (Series B announcement, 2024)
The funding announcement came alongside SewerAI's first seven-figure contracts - a milestone that signals the company has moved past land-and-expand pilots and into long-term municipal partnerships. The plan with Series B capital: expand engineering and go-to-market teams, improve AI training pipelines, and scale operations across North America.
From Touchscreens to Sewers
-
2002-03
Carnegie Mellon UniversityBS + MS in Electrical and Computer Engineering - a back-to-back degree from one of the world's top engineering programs.
-
2003-07
Fitbit - Senior Firmware EngineerHelped lead the team that created one of the first Android tablets, before Fitbit became synonymous with wristbands.
-
2007-10
Touch Revolution - Hardware Product ManagerOversaw development and production of consumer and industrial tablets and standalone touch panels.
-
2005-10
Centosette - Co-Founder & CTOBuilt data analysis and automation software for municipal water systems. Acquired by TPK - the world's largest touch panel manufacturer - in 2010.
-
2019
SewerAI - Co-Founder & CEOCo-founded with Billy Gilmartin to apply AI and computer vision to sewer inspection. What started as a weekend machine learning project became a Series B company.
-
2024
Series B & ScaleClosed $15M Series B led by Innovius Capital. SewerAI at 110 employees, 135M+ feet of inspection data under management, and first seven-figure contracts signed.
What Rosenthal Actually Says
"Most infrastructure was built post-WW2 and is reaching the end of its useful life, leading to more frequent failures and increased costs."
"Legacy incumbents offer on-premise or on-truck software that has seen very little innovation in the last 20 years."
"SewerAI's technology increases top and bottom lines by enabling more inspections per day at a lower cost."
"We're seeing an acceleration in demand for our platform as we enable people to do more with existing budgets."
What He's Built
- SewerAI AutoCode has processed over 100 million feet of sewer pipe inspections
- Raised $18.5M total funding, including $15M Series B closed June 2024
- Reduced sewer inspection costs by 40-70% for municipal clients
- Tripled inspection productivity for clients; up to 4,000 feet per day with AutoCode
- Secured major municipal clients including the City of Houston, San Francisco, and Toronto
- Saved tens of thousands of engineering hours and tens of millions of dollars for customers
- Closed first seven-figure contracts at SewerAI in 2024
- Doubled revenue year-over-year entering the Series B round
- Previously co-founded Centosette, acquired by TPK (world's largest touch panel manufacturer) in 2010
- Built SewerAI to 110 employees across engineering, field ops, and go-to-market teams
Things Worth Knowing
SewerAI's AI has watched more sewer footage than any human inspector alive - 135 million feet and still counting.
He holds both a BS and MS from Carnegie Mellon in Electrical and Computer Engineering, earned back to back.
SewerAI started as a weekend side project - Rosenthal was taking online machine learning courses when the idea clicked.
Centosette, his earlier startup, was acquired by TPK - the same company whose touchscreens appear in most of the world's smartphones.
Zachary Bookman, CEO of OpenGov, personally invested in SewerAI's Series B - a rare vote of confidence from civic tech's inner circle.
Before sewer AI, he helped build one of the first Android tablets at Fitbit. Different kind of screen time.