#StartupsEverywhere: Chip Kennedy, CEO, and Lindsay Avagliano, COO, CivicReach.AI
This profile is part of #StartupsEverywhere, an ongoing series highlighting startup leaders in ecosystems across the country. This interview has been edited for length, content, and clarity.
Improving government workflow with AI
With previous experience in the startup world, Chip Kennedy developed a passion for connecting humans and technology. From there, he founded CivicReach.AI, a platform using voice AI technology to improve customer experiences—especially at the local level—with the government. We sat down with Chip and Civic Reach.AI’s chief operating officer Lindsay Avagliano to discuss AI, working with governments, and more.
Tell us about your background. What led you to CivicReach.AI?
Chip Kennedy: I've been in the startup world my entire career, focusing on the different areas of technology that meet human experience. This ranges from travel experiences to years in healthcare. From there it seemed natural for me to transition into GovTech, specifically how the government can better serve people.
Lindsay Avagliano: I have previous experience working in government and economic development, specifically investing in startups in underserved regions of New York as part of their economic development agency. I've also worked for 15 years now in some combination of corporate law, venture capital law, venture capital investing, startup advisory work, and startup operations.
What is the work you all are doing at CivicReach.AI? Who are your main users and what problems do you solve for your customers?
Chip: The story of CivicReach is both long and short. We've been around for just over a year. In the very early days we spent time figuring out what it means to use voice AI to improve customer experience within government, especially at the local level.
The CivicReach platform is a customizable platform of voice AI with agents answering the phones on behalf of a government or agency. We are building an 'out of the box' virtualized 311 system which encompasses a municipality hotline, for a city, town, or county services.
We provide cities with an increased capacity for information delivery, when there are not enough administrators to pick up the phone or individuals are not doing their job. We also offer service delivery with AI agents that are agentic, meaning they can take actions to fill out forms, file requests, schedule services, and more.
We are bringing the AI world to the government, a space where workflows can be difficult and administrative help is needed. Finally, we provide governments with intelligence. One of the most trusted ways for residents communicating with their government is making phone calls. Governments want to have a mandate to provide for that line, but they don’t receive any data. In general, they are unaware of who is calling, the level of resident engagement, or the problems areas that are most discussed. We can use our software as a data source so the government knows more information about their resident callers.
Lindsay: It’s important to note, we are using a new frontier technology, but also governments are one of the least invested in when it comes to workflow innovation in general. We know our potential customers may not be prepared for technology to be trained on their existing workflow models. Though, with AI we no longer need to directly match the workflow step-by-step to the technology, we can supersede that.
How do you navigate the privacy, but also proprietary data concerns that local governments might have?
Chip: Data as a whole fuels AI systems, and ours is no different. We draw a line with personally identifiable information (PII), personal health information (PHI), and other sensitive data. In information and service delivery, we can do a lot without identifying information. We built our system so government entities have the option of deciding if we collect personal information from the user and if we keep it. The process is similar to an intern picking up the phone and filling out a form online. The intern isn’t trusted with personal information and transferring the information to a secure system; our AI works the same way.
Some governments ask us to store personal information and we're building our platform with that capability. One example of this might be storing data in a content management system of residents who call so service providers can have ideas of who needs utility bill assistance. Other governments want to save all transcripts, so we are working to build systems that can scrub any PII or PHI that accidentally comes into our system.
The beautiful thing about working with governments is that out of all the non-sensitive data industries, they have the most data by mandate, all the way down to local governments. Most information does become public, so we have a trove of data to pull from for training, improving, and delivering the product. For example, meeting minutes from a city council meeting went public directly following the meeting. We are simply ingesting it to give information to residents that they otherwise couldn’t access.
What role does AI play in your product?
Chip: We use state-of-the-art off-the-shelf large language models (LLMs) including those from Open AI, Anthropic, and Meta’s Llama. These models are our base layer, and our product is built to work with any model so that our customers can identify the model they’re most comfortable working with. Maybe an opinionated Chief Technology Officer has ideas, or the city is working with a model they’re already approved based on their internal AI regulation.
On top of that is our proprietary stack of the models we've built. We use a few different technologies and intellectual properties we built to create the effect of multi-agentic AI that sits in a virtual conference room.
When a caller calls in, they're still experiencing a phone call but there are human-sounding, AI agents sitting in on the call. Customers are alerted to the use of the AI agents, but we have found that they often forget because the job is done well. Additionally, other AI agents sit on the call to record that call and facilitate a warm transfer if the resident gets frustrated, asks to talk to a human, or asks to be passed to a specific department.
For each city government we work with, we build a mini data governance system tailored for them. Most governments are only recently taking steps to understand what data governance looks like, but they have a trove of data available. We create a system for them to have access to insights from said data pulled from city council minutes, websites, social media, and press releases. The information is ranked based on what is readily available and what residents are calling about. Then, the municipalities can tweak the formula in terms of what data comes in.
How do you approach IP as a startup founder? Why is a balanced attitude toward IP important for innovation?
Chip: As a technologist in the startup space, I've worked with a lot of entrepreneurs and founders who thought the best plan of action was to capture as much IP as possible. As a person building the code, that makes no sense.
There is a sort of unspoken agreement among all engineers that we borrow everything we build and most code is not our own. When working with software and software-enabled companies, it's clear how much is borrowed. It would be weird to capture intellectual property (IP) on how those systems come together. In general, we’ve been able to build our product because of open access to innovation and the idea that a lot of code is available to us.
In terms of our company, we are not building foundation models and LLMs. We are not changing how the underlying idea of generative AI could work. Instead, we’re taking those giant innovations and fitting them to an industry that is in desperate need of generative AI for a specific use case. So, we have IP on the truly proprietary things we have created that our competition hasn’t.
What are your goals for CivicReach.AI moving forward?
We are excited to be in a space that gets to serve government and public officials, especially local governments across the country. Oftentimes these are folks that work extremely hard, are incredibly underpaid, and choose careers that allow them to serve their communities with their skill set. The notion that we can provide tools for them to improve how they do their jobs, and serve their community more, is awesome. Not only is it a novel use of AI, but an incredibly important one.
As a company, we hope to do that for as many communities as possible by aggressively getting out there and telling people what we found we could do for them. Not only are we selling and working with governments, but we are also learning how to improve based on their feedback and make the product cost-effective.
In the long term, we want to make the idea of 311, which is often a policy conversation, into a technology conversation. The 911 system exists not only as technology but as a concept so that no matter where you are in the country, you can call and get emergency services. We think of this idea of 311 for all, which often is a policy conversation, but for the first time, we can make it a technology conversation.
All of the information in this profile was accurate at the date and time of publication.
Engine works to ensure that policymakers look for insight from the startup ecosystem when they are considering programs and legislation that affect entrepreneurs. Together, our voice is louder and more effective. Many of our lawmakers do not have first-hand experience with the country's thriving startup ecosystem, so it’s our job to amplify that perspective. To nominate a person, company, or organization to be featured in our #StartupsEverywhere series, email advocacy@engine.is.
