Ann Lewis
Washington, District of Columbia, United States
3K followers
500+ connections
3K followers
500+ connections
Activity
3K followers
-
Ann Lewis reposted thisAnn Lewis reposted thisToday I joined Mayor Mamdani, CTO Lisa Gelobter, and Maya Israni, to announce a historic investment in NYC’s capacity to deliver for New Yorkers - the PIT Crew 🏎️ 🏁 We’re building the NYC Public Interest Technology Crew, a team of senior software engineers, product designers, and product managers who will build better services for New Yorkers, faster and in-house. If we're serious about delivering on housing, childcare, consumer protection, and more, we need the capacity to make those promises real and felt faster. And that’s exactly what the PIT Crew is designed to do. Our team has a real mandate to build public products and services that make interacting with government dignified and delightful for New Yorkers. It’s an ambitious vision for the next era of civic tech, but we’re up for it! Huge thanks the The Rockefeller Foundation's NYC Affordability Lab and Danielle S. Goonan for their partnership and support from the very beginning. Thanks to DCWP Commissioner Samuel A.A. Levine for your leadership and for being our first partner agency. And a big shout-out to Sukhi Gulati-Gilbert, Mehgan Iulo, and Andrew Bunyea for stepping up to serve our city! Excited to be leading this team and for all the work ahead! 🏎️ 🏁
-
Ann Lewis shared thisThank you, Clare Martorana, for your partnership, leadership, and belief in what we’re building. You know what it takes to create meaningful public impact, and I’m thrilled to have your guidance as GroundVue grows. One thing you’ve consistently championed is that lasting government transformation isn’t about technology alone—it’s about people, institutions, and the hard work of turning possibility into better outcomes. We’re grateful to have you on this journey with us.Ann Lewis shared thisI just joined GroundVue as an advisor. Why? The future of government won’t be built by technology alone. It will be built by people who know how to turn technology into public impact. I've worked with many talented leaders across government, technology, and the public sector. What stands out about GroundVue's founders is that they have spent their careers not just talking about change, but delivering it. Shannon Arvizu, Ph.D. advised the White House Chief Data Scientist, led data strategy for a $12B federal agency, and founded Counties R.I.S.E. at University of California, Berkeley. She understands how data can drive better decisions. Ann Lewis led the Technology Transformation Services (TTS) at GSA, building products and platforms that improved how government serves the public. She knows what it takes to transform institutions from the inside. Travis Hoppe served as the Centers for Disease Control and Prevention's first Chief AI Officer and as Assistant Director for AI R&D at the The White House Office of Science and Technology Policy. He has helped shape the federal government's approach to AI at a pivotal moment for the country. As AI is rapidly changing what's possible, the real challenge isn't adopting new technology, it is ensuring that technology creates meaningful public value. This team understands that. For those focused on impact, resilience, and long-term value creation, few opportunities rival helping governments deliver better outcomes for the people they serve. Contact Shannon Arvizu, Ph.D. for more information. GroundVue #CivicTech #AI #PublicSector
-
Ann Lewis reposted thisAnn Lewis reposted thisI just joined GroundVue as an advisor. Why? The future of government won’t be built by technology alone. It will be built by people who know how to turn technology into public impact. I've worked with many talented leaders across government, technology, and the public sector. What stands out about GroundVue's founders is that they have spent their careers not just talking about change, but delivering it. Shannon Arvizu, Ph.D. advised the White House Chief Data Scientist, led data strategy for a $12B federal agency, and founded Counties R.I.S.E. at University of California, Berkeley. She understands how data can drive better decisions. Ann Lewis led the Technology Transformation Services (TTS) at GSA, building products and platforms that improved how government serves the public. She knows what it takes to transform institutions from the inside. Travis Hoppe served as the Centers for Disease Control and Prevention's first Chief AI Officer and as Assistant Director for AI R&D at the The White House Office of Science and Technology Policy. He has helped shape the federal government's approach to AI at a pivotal moment for the country. As AI is rapidly changing what's possible, the real challenge isn't adopting new technology, it is ensuring that technology creates meaningful public value. This team understands that. For those focused on impact, resilience, and long-term value creation, few opportunities rival helping governments deliver better outcomes for the people they serve. Contact Shannon Arvizu, Ph.D. for more information. GroundVue #CivicTech #AI #PublicSector
-
Ann Lewis reposted thisThe #GroundVue Team is doing something special - a quick read that will leave you inspired. Congrats Ann Lewis, Travis Hoppe & Shannon Arvizu, Ph.D. #AIforGood #OpenAI #civictechAnn Lewis reposted thisCodex is helping Travis Hoppe, Ann Lewis, and Shannon Arvizu, Ph.D. build a new way for governments to learn from one another. Their company, GroundVue, starts with a problem local officials know well: cities, counties, and states are often working through the same hard challenges, from homelessness to infrastructure to public health, without an easy way to see what other governments have already tried. The answers often exist. They are just buried inside public meetings. GroundVue turns those meetings into information officials can search and compare. A mayor looking for ways to prevent homelessness could see how other cities are helping families stay in their homes before eviction. A county confronting an unfamiliar challenge could learn what similar places debated, decided, and discovered. “We’re now creating a way to make government searchable in real time,” said Shannon, a sociologist and policy researcher. GroundVue is solving a difficult data problem. Across the US, roughly 90,000 government bodies make decisions that touch people’s lives. About half of their public meeting videos are available in obvious places, like YouTube. The rest are spread across local websites and other platforms, with no single format or easy way to find them. GroundVue uses Codex to help find those harder-to-reach public sources and build the systems needed to keep collecting them. OpenAI models help transcribe the meetings and identify the problems, tradeoffs, and possible solutions officials are discussing. For Ann, who previously led Technology Transformation Services at the General Services Administration, the change in speed is dramatic: “All those steps that previously took a few days or a few weeks take us literally minutes now.” Travis joined the company after serving as the chief AI officer at the Centers for Disease Control and Prevention, where he used OpenAI’s Deep Research in a pilot program to help the agency process large volumes of research and data. GroundVue reflects a related idea: AI can help public institutions make sense of information that would otherwise be too scattered and time-consuming to use. The analytical work has changed, too. “Before AI, this would have taken an army of technologists and an army of social scientists,” Shannon said. GroundVue is still in its infancy. The company plans to launch first in Virginia, California, and Minnesota over the next few months, expand nationally later this year, and add full historical data to the platform within two years. The point is not to replace public servants or local judgment. It is to help officials learn from the decisions and debates already taking place in communities across the country. As Travis puts it, “Instead of using generative AI to generate things, we can use it to listen to the things that people are actually saying.”
-
Ann Lewis reposted thisAnn Lewis reposted this🤯✨ We are directly featured in OpenAI’s new report, “The Next Era of Knowledge Work,” alongside examples of how AI is changing what small teams can build. And yes, I am going to be excited about this for a minute. We are a small startup. OpenAI is, well, OpenAI. Having them recognize the work we are doing feels meaningful because it points to something much bigger than us. We are building civic data infrastructure for a new kind of knowledge work, turning fragmented public information into structured intelligence. Public decisions shape our communities every day, but the information behind them is often difficult to access. That is the gap we are closing. A big reason we have been able to move so quickly is Codex. GroundVue operates in an environment with thousands of fragmented data sources, constantly changing websites, and highly variable formats. Historically, building and maintaining systems to discover, collect, normalize, and organize this information would have required a much larger engineering team. Coding tools like Codex change that equation. It allows us to automate significant portions of the work involved in building and maintaining data pipelines, exploring new sources, testing approaches, rapidly iterating on infrastructure, and developing the harness engineering needed to reliably evaluate, monitor, and improve AI systems in production. Tasks that once took days or weeks can often be completed in hours, allowing a small team to operate with a level of leverage that would have been difficult to imagine just a few years ago. That leverage is not just about productivity. It is what enables us to continuously find hard-to-reach public sources, organize distributed information, and transform it into structured policy intelligence that can be used by public servants, researchers, nonprofits, journalists, and communities. It is agentic workflows at their finest. Small teams can now build infrastructure that previously required far greater resources. By combining AI, data, tools, and workflows into coordinated systems, we can create entirely new forms of infrastructure that help people understand public decision-making at a scale that was previously out of reach. https://lnkd.in/eKANmSTp #AI #CivicTech #Codex
-
Ann Lewis reposted thisAnn Lewis reposted this🤯✨ We are directly featured in OpenAI’s new report, “The Next Era of Knowledge Work,” alongside examples of how AI is changing what small teams can build. And yes, I am going to be excited about this for a minute. We are a small startup. OpenAI is, well, OpenAI. Having them recognize the work we are doing feels meaningful because it points to something much bigger than us. We are building civic data infrastructure for a new kind of knowledge work, turning fragmented public information into structured intelligence. Public decisions shape our communities every day, but the information behind them is often difficult to access. That is the gap we are closing. A big reason we have been able to move so quickly is Codex. GroundVue operates in an environment with thousands of fragmented data sources, constantly changing websites, and highly variable formats. Historically, building and maintaining systems to discover, collect, normalize, and organize this information would have required a much larger engineering team. Coding tools like Codex change that equation. It allows us to automate significant portions of the work involved in building and maintaining data pipelines, exploring new sources, testing approaches, rapidly iterating on infrastructure, and developing the harness engineering needed to reliably evaluate, monitor, and improve AI systems in production. Tasks that once took days or weeks can often be completed in hours, allowing a small team to operate with a level of leverage that would have been difficult to imagine just a few years ago. That leverage is not just about productivity. It is what enables us to continuously find hard-to-reach public sources, organize distributed information, and transform it into structured policy intelligence that can be used by public servants, researchers, nonprofits, journalists, and communities. It is agentic workflows at their finest. Small teams can now build infrastructure that previously required far greater resources. By combining AI, data, tools, and workflows into coordinated systems, we can create entirely new forms of infrastructure that help people understand public decision-making at a scale that was previously out of reach. https://lnkd.in/eKANmSTp #AI #CivicTech #Codex
-
Ann Lewis reposted thisAnn Lewis reposted thisNew book alert: Eric Ries just dropped a bomb with Incorruptible. Eric has long been the “go to” voice on how companies build, test, and scale with the Lean Startup approach. Now he is saying something every founder and investor should take seriously: It is time for business to change. His argument is direct: • Good companies do not go bad because bad people take over. They go bad because their structures and incentives slowly pull them away from their purpose. • Mission cannot just live in a pitch deck. It has to be built into the legal structure, capital model, board design, and operating system of the company. • Governance is not bureaucracy. Done right, it is one of the most creative acts of company-building. This is exactly why we formed GroundVue as a Public Benefit Corporation. Our legal charter codifies our purpose: Build enduring civic data infrastructure that "informs the design, monitoring, and implementation of public policies that directionally improve household economic security and quality of life." That purpose is not a side note. It is part of the company’s foundation. It shapes how we use AI responsibly. It shapes how we partner with mission-aligned funders. It shapes how we uphold credibility with public sector customers. And it shapes how we think about equitable access, so we serve any local government of any size. Admittedly, it's not easy to walk the talk. This is still new language for most funders. As of last week, however, I can now say: "Don't take our word for it. Just read the book."
-
Ann Lewis shared thisExcited to see GroundVue featured in Second Opinion Media alongside other companies exploring how AI can strengthen public-sector capacity. One of our core beliefs is that some of the most important data in America is hiding in plain sight: public meetings, local decisions, implementation discussions, and operational signals that are incredibly difficult to analyze at scale. We’re using AI to help transform that fragmented information into usable civic intelligence — helping governments, nonprofits, and communities learn faster, compare approaches, and respond to emerging challenges in real time. Thanks to Ruth Reader for the thoughtful piece. cc Travis Hoppe Shannon Arvizu, Ph.D.Ann Lewis shared thisBig companies, startups, and health systems are all thinking about how to do more with less and using AI to get more efficient. That means health systems big and small are buying fewer products. Savvy startups are raising less money and getting strategic about who they're selling to. One emerging group is thinking about how they can use AI to access untapped or difficult to get at health data and selling their tools to governments, federally funded clinics, and other less considered entities. For this I chatted with CDC's former head of AI, Travis Hoppe, as well as his cofounders Shannon Arvizu, Ph.D., and Ann Lewis about their new startup, Groundvue, and the many opportunities to apply AI to unstructured public datasets. https://lnkd.in/epQtyTbC
-
Ann Lewis reposted thisAnn Lewis reposted thisI will not shut up about MCPs. Sorry in advance! 📣 I was grateful to join a panel at Red Hat Summit on “Trusted intelligence: Building the next generation of public health systems with MCP servers,” alongside experts Susan Gregurick / NIH and Benjamin Cushing / Red Hat. We discussed the growing opportunity to use the Model Context Protocols (MCPs) to improve how public health systems access, interpret, and act on data across complex environments. Public health infrastructure faces many of the same challenges that come up across federal data systems: fragmented data sources, uneven access patterns, jurisdictional complexity, governance requirements, privacy constraints, and the need for reliable provenance and explainability. MCPs offers a practical path forward. It gives AI assistants and agentic systems a standard way to connect with data sources, analytic tools, and operational workflows while preserving the trust and control that public institutions require. That matters for use cases like federated disease surveillance, outbreak response, privacy-preserving data integration, and real-time analytics. It also matters for the broader goal of lowering barriers to public data so people can use it in meaningful ways: informing decisions, building tools, shaping policy, and solving real problems in their communities. Go check out the work from other Federal leaders like Bella Mendoza Haley Johnson and Mark Aronson, who are working on it in real-time. This is also closely connected to the work we are building at GroundVue. Think how amazing it would be to have an infrastructure of civic data! #AI #MCP #AgenticAI #PublicHealth
Ann Lewis commented on a post
2mo
It’s an interesting point that many public systems implicitly rely on friction as a form of demand management. But with or without “agentic flooding,” most government services still lack transparency into what capacity their systems can actually handle. Agencies are often surprised by entirely organic demand spikes, like call centers overwhelmed during the pandemic or outages cascading across dependent systems. I think the deeper issue is that many government operating models were not designed for the level of responsiveness or observability we expect from modern consumer systems. Both organic demand and potential agentic demand point to the need for basic operational visibility across government: uptime, availability, throughput, queue depth, processing times, and service-level expectations. Some agencies do this well, but across government these metrics are often missing or difficult to discover.
-
Ann Lewis liked thisAnn Lewis liked thisWe've shipped an initial MCP server for Daisychain that allows customers to bring information about their supporters into Claude or ChatGPT, or send text messages to volunteers from your AI chats. https://lnkd.in/ghTxnqXT There are of course appropriate organizational controls and granular permissions are enforced. Let us know if there are more MCP Tools you'd like us to expose & what use cases you might have, but we have a bunch planned building on top of this initial minimalist surface.
-
Ann Lewis liked thisAnn Lewis liked thisBuilding and leading the Maryland Digital Service (MDDS) has been an incredible honor and one of the highlights of my career. This is a dream team of 63 exceptional public servants who show up every day with curiosity, build trusted relationships across agencies, and focus on 'fixing all the things.' This is the first time in the state's history that FT employees have represented such a broad range of modern digital skills. It takes a unique group of people to run toward the challenges of building something new in government, and I have been constantly inspired by their relentless dedication to transforming Maryland's digital services. I am immensely proud of what we have accomplished together: - Platform Modernization: Rebuilding hundreds of websites onto a secure, accessible, and reliable new platform (a journey that continues!). - Enterprise Tools: Developing the Maryland Web Design System (that and more at digital.maryland.gov) and analytics.maryland.gov. - Procurement Options: Creating and awarding a $300M IDIQ contract vehicle to enable scaled procurement change across the state. A significant focus for MDDS was improving the delivery of our Major IT projects. This is a large portfolio (60+ projects) representing hundreds of millions of dollars in investment. To get better outcomes, we needed a fundamental shift in how Maryland approaches technology. Through both oversight reform and hands-on delivery support, we helped agencies shift from a mindset of technology implementation to one of service delivery. Shifting to outcome-driven, user-centered, and iterative delivery sounds easy and intuitive—but this change is incredibly hard. The status quo is deeply reinforced by a long-held focus on compliance over outcomes. Technology alone will not solve these problems; lasting progress requires deep partnership across procurement, hiring, budget, legislative bodies, and agency leaders. We have proven that change is possible. I am incredibly inspired by our success across the projects we supported—creating alignment on the actual problems to be solved, bringing constituent and staff voices into the discovery process, and shifting projects toward iterative, "ship-and-learn" behaviors. Sometimes, a journey reaches a point where it needs a fresh perspective. I believe my departure will open up space for fresh ideas and leadership to push this work forward in ways. I am excited to see where the next leader takes this incredible group of people. To my team: Thank you for your dedication, your brilliance, and your support. Thank you for being the founding members of MDDS. I will be cheering you on from afar. To my partners across MD agencies, DoIT, and the Governor’s Office: Thank you for your collaboration and support as we built and flew the plane at the same time. Katie Savage, thank you for the opportunity and support to create this special organization. Jennifer Smith David HolmesLilly MadiganLauren GilchristSuran De SilvaClare Bayley
-
Ann Lewis liked thisAnn Lewis liked thisGovernments are spending billions solving problems that other governments have already cracked. The IMF estimates countries could gain 30-40% more value for money simply by adopting practices already proven to work elsewhere. Government AI Navigator, built by Apolitical, is designed to solve this. It's a living library of hundreds of real government AI projects from over 50 countries, to help public servants learn from what's already been tried. Take a look: https://lnkd.in/gMeWXgaW
-
Ann Lewis liked thisAnn Lewis liked thisI have joined the Center for American Progress as a Senior Fellow for Government Delivery, working on an ambitious service delivery and reform agenda. More on that later. But in the meantime, thrilled to support the new NYC PIT Crews from CAP and we just issued this statement: “This ambitious move shows that New York City’s leaders are committed to building responsive, easy-to-use government services. Hopefully, this will be a model for cities, counties, and states across the country. I encourage them to tackle the impossible projects; nobody thought free tax filing was possible, but an in-house government team in the Biden administration made Direct File happen. They should work to set high standards for improving government efficiency.” Congrats again to Lisa Gelobter Luke Farrell Maya Israni and team 👏 https://lnkd.in/gNzxFMJrSTATEMENT: CAP’s Merici Vinton Praises New York City’s Public Interest Technology CrewSTATEMENT: CAP’s Merici Vinton Praises New York City’s Public Interest Technology Crew
-
Ann Lewis reacted on thisAnn Lewis reacted on thisToday I joined Mayor Mamdani, CTO Lisa Gelobter, and Maya Israni, to announce a historic investment in NYC’s capacity to deliver for New Yorkers - the PIT Crew 🏎️ 🏁 We’re building the NYC Public Interest Technology Crew, a team of senior software engineers, product designers, and product managers who will build better services for New Yorkers, faster and in-house. If we're serious about delivering on housing, childcare, consumer protection, and more, we need the capacity to make those promises real and felt faster. And that’s exactly what the PIT Crew is designed to do. Our team has a real mandate to build public products and services that make interacting with government dignified and delightful for New Yorkers. It’s an ambitious vision for the next era of civic tech, but we’re up for it! Huge thanks the The Rockefeller Foundation's NYC Affordability Lab and Danielle S. Goonan for their partnership and support from the very beginning. Thanks to DCWP Commissioner Samuel A.A. Levine for your leadership and for being our first partner agency. And a big shout-out to Sukhi Gulati-Gilbert, Mehgan Iulo, and Andrew Bunyea for stepping up to serve our city! Excited to be leading this team and for all the work ahead! 🏎️ 🏁
-
Ann Lewis liked thisAnn Lewis liked thisCongrats to Lisa Gelobter, Luke Farrell, Maya Israni for the launch of the incredible PIT (public interest technology) crew. First stop: Click-to-cancel protections to crack down on junk fees and subscription traps, and make it easier for people to know the real price of what they are buying and to stop paying for the services they no longer want. PIT Crew: https://nyc.gov/pitcrew + announcement: https://lnkd.in/eW3cMSnr Click-to-Cancel Rule: https://lnkd.in/e8xbtKyA
-
Ann Lewis liked thisAnn Lewis liked thisNYC goes big on digital capacity 🗽 - today, Mayor Mamdani, CTO Lisa Gelobter announced their PIT Crew. This is an in house team, supporting services across the city, committed to producing tangible outcomes, and building capacity for good outcomes, not just when something bad happens. This is a proactive, engaged vision for governing, not just reactive. My advice to this team - go big or go home - by building this incredibly talented team, there is nothing too big you can't do. Find the impossible projects - everyone said free tax filing was impossible, but an in house team made it happen. Set incredibly high standards, and deliver for New Yorkers. Build open source, public options that can be used by and a model for the rest of the country. 🗽 Watch and read the announcement here - this team is doing the right thing and announcing their first project, Click to Cancel, demonstrating impact on day one. Huge congrats to Lisa, Luke Farrell, Maya Israni, Sarah Milstein and team 🗽 https://lnkd.in/ggYqhhe7Mamdani forms Public Interest Technology crews to improve government efficiency and affordabilityMamdani forms Public Interest Technology crews to improve government efficiency and affordability
Education
-
Carnegie Mellon University
-
-
Activities and Societies: Winner of the John S. Morrison scholarship, covering 100% of tuition, room/board.
Winner of the InformationWeek Summer Fellowship for the project: Implementing a provably secure steganographic text encoding module for Emacs.
Relevant coursework: computer networks, algorithms, machine learning, artificial intelligence, computer music, computer systems, data structures, web server programming.
Paper: A Probabilistic Model of Melodic Similarity, http://citeseer.ist.psu.edu/590349.html
Projects
-
Rosetta Stone Advanced English for Business
SproutCore and HTML5 Web application, backed by a NoSQL database, that provides advanced business English training through videos, matching, sequencing, fill-in-the-blank, find-the-error, speech recognition, expert writing or speaking review, and live coaching sessions.
Other creators -
-
Rosetta Stone Advanced English for Business
SproutCore and HTML5 Web application, backed by a NoSQL database, that provides advanced business English training through videos, matching, sequencing, fill-in-the-blank, find-the-error, speech recognition, expert writing or speaking review, and live coaching sessions.
Other creators -
Recommendations received
19 people have recommended Ann
Join now to viewOther similar profiles
Explore more posts
-
Erika Harrison
PSA: The age where developers smile and nod - or run - when conversations include "I have an idea for an app..." is over. It still takes time to put together applications, but the amount required is So. Much. Less. Especially when blending different platforms, 3rd party APIs, large codebases, etc. If you're not using AI, you're not doing your job. No excuses. 💪 I anticipate companies that will survive and thrive in the coming years will be those that provide good support, user delight, and of course do what customers need them to do.
1 Comment -
Abe Gong
Spec-driven development has roughly doubled my coding output over the last couple months. It's also created a ton of documentation exhaust. Most discussion about spec-driven workflows ends at "ship the feature." Nobody talks about what to do with the spec after. I've iterated several times on this — from deleting specs immediately, to keeping a graveyard folder, to a new approach I'm pretty excited about: graduating spec content into permanent documentation when work ships. I wrote up a post covering what we learned through all three iterations, plus the templates and skills we're using in production today. Link in the first comment.
9 Comments -
Dasith Wijesiriwardena
I’ve been working on the #AAuth .NET SDK (still in draft), and wrote a blog post to explain the IAM problem space around agentic delegation in a simpler way for folks new to the domain. Agents are easy to talk about in abstract terms: “let the AI do things for me.” But the moment an agent touches real accounts, APIs, money, messages, or data, the hard questions show up: who is the agent, who is it acting for, what exactly was it allowed to do, and how do we know when that authority should stop? The post uses a parent sending a kid to the corner shop (90s themed) on the family account as the analogy. It maps surprisingly well to agent identity, delegated authority, consent, missions, revocation, sub-agents, call chains, and audit. AAuth is used as the reference model, but this is more of an informal introduction to the problem domain. Sharing in case others find the framing useful. Slides: https://lnkd.in/gRF_Dk_x Post: https://lnkd.in/gF3PTGyJ
2 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content