Frequently Asked Questions

CityData.AI is a big data and geospatial AI company that creates dynamic urban digital twins for smarter, sustainable, and resilient cities. We specialize in mobility intelligence, including people counting, density patterns, and movement trends for cities, business districts, parks, and public spaces, helping local governments make data-driven decisions across parks and recreation, economic development, transportation, public works, and community engagement.

CityData.AI is headquartered in San Francisco, California, at 548 Market Street, San Francisco, CA 94104. The company also has registered offices in Canada, Mexico, India, and Singapore.

Yes. CityData, Inc., operating as CityData.AI, is a US C-Corporation incorporated in the State of Delaware, with primary business operations in the United States.

Yes. CityData.AI is a profitable, self-sustaining company. After a modest seed round, we have grown through fiscal discipline and a relentless focus on our civic customers, and we have been profitable for the past several years.

Our coverage spans the United States, Canada, Mexico, and major regions in South America, Europe, and Asia, and we can activate urban digital twins for virtually any municipality in the world. Since 2020, our customers have included cities, academic institutions, and research organizations across North and South America, Europe, the Middle East, Asia-Pacific, Australia, and New Zealand.

CityData.AI was founded in October 2020 by Apurva Kumar, who serves as CEO and acting CTO. Apurva holds a Master of Science in Engineering and Earth Sciences from Stanford University and brings decades of experience in geostatistics, data science, mobile analytics, machine learning, and large-scale geospatial platforms.

Our team members are alums of institutions such as Stanford University, USC, UC Irvine, IIT, and ETH Zurich, with deep expertise in mobility data intelligence, AI/ML, geospatial analysis, large-scale data compute, and cloud infrastructure. You can learn more at citydata.ai/company.

Each product is fine-tuned for a specific government agency's needs. CityParks serves Parks and Recreation, CityEconomy serves Economic Development and Urban Planning, CityFlow serves Transportation agencies, CityOps supports Public Works and field operations, and CityChat focuses on community engagement.

CityData.AI offers a comprehensive portfolio of AI products for communities: CityFlow (mobility and transportation), CitySim (scenario planning and simulations), CityEvents (event attendance analytics), CityEconomy (spending and economic development), CityParks (parks and recreation visitation), CitySensor (solar-powered optical counting sensors), CitySurvey (mobile ground-truth surveying), CityOps (field operations and workforce management), CityChat (the Cici agentic AI chatbot), and CityBase (reservations, registrations, and bookings).

CityParks replaces outdated manual counting with precision analytics for parks and recreation departments. It combines privacy-compliant mobility data with ground-truth sensor calibration to deliver accurate visitation counts, movement trends, dwell times, hourly and seasonal patterns, weather correlations, and visitor origin and demographic insights for parks, trails, facilities, and event venues.

CityFlow is our mobility intelligence product for transportation planners. It provides origin-destination matrices, trip trajectories, multimodal split analysis (driving, walking, cycling, transit), and segment-level traffic metrics, helping departments of transportation understand travel demand, visualize congestion, and plan more efficient routes and infrastructure.

CitySim is our agent-based modeling and simulation engine for transportation engineers and urban planners. It generates high-fidelity synthetic populations from real-world mobility data, census demographics, and travel surveys, letting you test 'what-if' infrastructure scenarios, predict traffic outcomes, and validate interventions before committing real-world resources.

Yes. CityEconomy analyzes foot traffic in downtowns and commercial corridors to help economic development organizations measure district vitality, prove value to prospective retailers, track downtown recovery, understand trade areas and cross-visitation, and quantify the economic impact of tourism and events.

CityEvents estimates attendance, movement trends, demographics, before-and-after behaviors, and spending patterns for events, festivals, recreational programs, and sporting and entertainment activities at stadiums, arenas, convention centers, fairgrounds, parks, and business districts.

CitySensor is a privacy-compliant, ruggedized IoT sensor that municipal agencies can mount on streetlight poles, parking lots, parks, trailheads, or community centers to accurately count people, vehicles, and animals in real time. It transmits count data over 4G, 5G, or WiFi to your cloud dashboard and serves as ground truth to validate and calibrate our AI models.

CitySurvey is a mobile app for iOS and Android that uses computer vision AI to count and survey people, bicycles, animals, and vehicles. It can run continuous autonomous surveys from a phone on a tripod at parks, trailheads, stadiums, and festivals, with results relayed to a cloud dashboard and used as ground truth for our visitation models.

CityOps is an AI-powered command center for mobile workforces. Through a privacy-first, opt-in network, it connects managers to field crews with location-aware notifications and workflows, route tracking, time-on-site measurement, and exports to existing task management and payroll systems.

CityChat connects residents to their local government through 'Cici,' a multilingual, 24/7 AI digital concierge. It surfaces hyper-local updates such as 311 reports, construction permits, community events, and park details, and drives participation through localized surveys, polls, and neighborhood alerts on iOS, Android, and the web.

We model population movement during crises to identify evacuation bottlenecks, measure population displacement during floods or fires, and help emergency managers plan resilient infrastructure based on actual human behavior. Our simulation capabilities have supported resiliency projects including work with United Nations SIDS programs.

Yes. CitySim provides agent-based simulations that help cities plan for urban air mobility (eVTOLs and drones), analyzing ground traffic and commuter demand to identify optimal vertiport locations and model the impact of air traffic.

No. Our platform is location-agnostic and can analyze any geofenced area, including downtown zones, business improvement districts, retail corridors, industrial zones, trade areas, event venues, and mixed-use developments, in addition to parks, trails, green spaces, campgrounds, water bodies, and natural resources.

Simply fill out our online form for a free consultation. We will assign you a dedicated customer success manager and account manager, work with you to define the parks, districts, venues, or networks you want to analyze, activate your account, and configure your dedicated dashboard.

You can provide standard GIS files (Shapefiles, KML, or GeoJSON) outlining your boundaries, draw shapes yourself using tools like Google My Maps, Google Earth, or ArcGIS, or simply send us a list of names or addresses and our geospatial team will map them for you as a free courtesy service.

Yes. Geofences can be modified if boundaries change or you want to refine your analysis. Minor modifications are typically included in your subscription; extensive re-mapping or large-scale changes that require significant historical data reprocessing may incur additional fees depending on your contract.

Yes, submitting your own official shapes is the preferred method. It ensures the analysis matches your department's exact boundary definitions, including or excluding features like parking lots or maintenance yards as desired.

Shape ingestion runs through a data pipeline that includes cleaning, calibration, historical backfilling, and ground truthing. Generating the full historical dataset for new geofences typically takes 3 to 5 business days.

No. Our scalable cloud infrastructure supports unlimited places and geofences, whether you need to analyze five parks or hundreds of thousands of road segments, without performance degradation.

Most of our products are turnkey, and we typically follow a structured 5-to-6-week implementation plan from contract signing: weeks 1-2 cover kickoff, geofencing, and data ingestion; weeks 3-4 cover computation, optional sensor deployment, and an initial dashboard review; and weeks 5-6 cover refinement, final delivery, training, and publishing your branded dashboard.

No. CityData.AI provides unlimited user licensing for your organization. You can grant access to as many team members, stakeholders, and inter-departmental colleagues as needed without paying for extra seats.

No. CityData.AI is a fully cloud-native, web-based platform accessible from any modern browser (Chrome, Edge, Firefox, Safari) on any device with an internet connection.

Our insights are derived from a massive pool of anonymized, privacy-compliant mobile location signals sourced from mobile apps and SDKs with explicit user consent, along with connected vehicle and IoT sensor feeds. The raw data is cleaned, aggregated, and processed through proprietary machine learning algorithms to represent population-wide movement trends.

No. For first-party sources we are contractually obligated to withhold app identities, and third-party aggregators typically strip app-specific identifiers before data reaches us. This keeps our analysis neutral and focused on aggregate movement rather than behavior within any specific app.

We never extract demographics from individuals. Instead, we identify the anonymized common evening location (presumed home area) of a device at the Census Block Group level and apply aggregate census characteristics of that area, such as median income and age distribution, to the visitor profile.

By default, a device must remain within a geofence for at least 4 minutes to register a visit. This threshold is fully configurable: we recommend around 10 minutes for regional or state parks, and category-appropriate thresholds for commercial locations such as cafes versus shopping malls.

We primarily provide high-fidelity historical data with a core dataset latency of 2-3 days, refreshed on a monthly cadence for maximum accuracy. Event-focused insights can be delivered within 7-14 days, and near-real-time counting is available through our CitySensor hardware.

A monthly cadence aligns with standard urban planning and reporting cycles, ensures a statistically robust sample size, and allows time to clean, calibrate, and verify the data against ground truth. Weekly splits can be volatile due to weather and short-term anomalies.

Data for the previous month is processed, quality-checked, and published to your dashboard by approximately the 10th day of the current month.

Our historical archives generally extend back to January 2019 or 2020, depending on the region. This deep history supports longitudinal studies, such as comparing current visitation against pre-pandemic baselines.

Yes. We capture visitation spikes associated with festivals, concerts, and public gatherings within your geofences, so you can measure the impact and reach of specific community events. Event-specific reports can typically be delivered within 7 to 14 business days after an event.

Yes, it is highly recommended. Sharing event dates allows our data scientists to contextualize anomalies, validate peak counts, and generate Event Impact Reports that isolate event traffic from normal baseline usage.

Yes, and we are transparent about them. We do not collect data from children, and mobile data from seniors (65+) is reduced due to lower app usage in that cohort. Their presence is statistically inferred using census household composition data, allowing us to estimate family and senior visitation without any digital surveillance of those groups.

Visits are computed from crowdsourced mobility data and scaled with an expansion ratio from our gravity AI model, while trip routes and modes are based on interpolated raw data. Trips are typically not scaled up in CityParks and CityEconomy dashboards, though they are scaled up in CityFlow and CitySim.

Heatmaps show the relative spatial density of movement within a boundary, revealing where visitors cluster, such as entrances, playgrounds, or plazas. Visitation charts measure visit volume over time, while heatmaps show where activity concentrates, so differences between the two are expected and do not indicate data inconsistencies.

Heatmaps use pre-aggregated, anonymized GPS density data and are designed to show stable overall spatial patterns rather than short-term fluctuations. A quarterly filter aggregates movement into three-month periods to reduce noise and provide representative seasonal patterns.

Ground truth is real-world verification data collected on-site, such as manual counts, surveys, sensor readings, or ticket sales. We use it to benchmark our digital mobility signals and ensure our AI models accurately reflect reality.

Mobile data penetration varies by region and demographic. Comparing our digital signal counts against physical ground truth counts lets us calculate a precise expansion factor to scale our data to the total population with high confidence.

We recommend at least one week of ground truth data per month, covering both weekdays and weekends and ideally different times of day. A simple CSV or Excel file with Date, Time, and Count columns is the most effective format.

We use a ratio-based calibration method: we calculate the correlation between mobile devices observed in a zone and the actual people counted on the ground, then apply that expansion factor dynamically to historical and future data to derive total visitation estimates.

Yes. We actively collect and curate ground-truth data from open sources wherever available, including transit ridership counts, highway entry/exit counts, airport visitation, road surveys, event attendance, parks and trails counters, tourism board counts, thermal satellite imagery where available, and US Census population data.

We apply strict, automated quality control to all raw mobility data. Our algorithms continuously scan for anomalies, hardware issues, unrealistic movement patterns, spoofed locations, implausible speeds, and fraudulent activity, categorizing every observation with detailed quality and accuracy flags before it enters our models.

Yes. CitySensor is a proprietary, ruggedized device for permanent outdoor installation that uses edge-computing computer vision AI to count pedestrians, cyclists, vehicles, and animals continuously in real time, with results automatically uploaded to our cloud platform.

CitySensor is flexible: it can be hardwired into mains power, such as on a light pole, or deployed as a low-power version with solar panels and battery backup for remote trail locations.

No. CitySensor uses computer vision to detect the shape and movement of a person or bicycle but does not capture, store, or analyze facial features or any biometric data. All image processing happens on the device, no video is recorded or transmitted, and only numerical counts are sent to the cloud.

Yes. The CitySurvey app is free to download and use for CityData.AI customers on both iOS and Android. It replaces the traditional clicker for manual counts and helps mobilize staff or volunteers for counting and ground-truthing projects.

Yes. Data collected via the CitySurvey app or CitySensor hardware is treated as the gold standard of ground truth and is prioritized in our calibration pipeline to refine expansion ratios and fine-tune the accuracy of our city-wide mobility models.

CityFlow uses location signals typically accurate to 1-20 meters, with spatiotemporal clustering requiring multiple data points to confirm a trip segment. Data penetration typically ranges from 35% to 65% across road geometries, and machine-learning scale-up techniques calibrated against census data, traffic counters, toll data, and other ground truth align sample counts with true regional volumes.

Yes. Because our movement data is a statistically significant sample rather than 100% of movement, we calculate dynamic scale-up multipliers for each recorded trip by correlating our sample against ground-truth and census datasets. These multipliers are critical for accuracy when feeding our data into agent-based simulators, including CitySim.

Yes, completely. Data is stripped of all personal identifiers such as names, phone numbers, and email addresses before it is ever processed, and it is aggregated so that it only reveals visitation counts and movement patterns for groups, never for an individual.

No. We do not track individuals or build profiles on specific people. Our system focuses exclusively on aggregate patterns, processing billions of data points to understand how groups of people move through a city or park.

We follow Privacy by Design principles and never process personally identifiable information. We hash mobile device identifiers, blur GPS coordinates, aggregate data into grid cells and time buckets, apply differential privacy and noise techniques, and enforce k-anonymity thresholds so no individual can be isolated in the dataset.

No. We maintain a strict policy against collecting data from children and do not knowingly collect, process, or store data from devices associated with individuals under 18. Family visitation is estimated statistically using census household composition data, not by tracking minors' devices.

Yes. Our data activities comply fully with applicable laws, including US state consumer privacy laws (such as CCPA and CalOPPA), GDPR (EU), PDPA (Singapore), and LGPD (Brazil). We are a registered data broker in the State of California, listed on the Attorney General's registry, and we continuously monitor and comply with evolving state legislation such as Oregon's privacy laws and Maryland's MODPA.

No. Unlike companies that monetize citizen data, CityData.AI strictly does not collect, provide, or sell citizen data for advertising purposes to brands, retail chains, shopping malls, or commercial real estate owners. Our products are built exclusively for government use cases.

Yes. Mobile publishers who partner with CityData are required to implement strict opt-ins and opt-outs to obtain affirmative consent from end users before any data collection.

Yes. Datasets and insights are strictly siloed with granular access controls, so data is only accessible to authorized city staff and essential CityData support personnel. Reports created by city staff are never shared with other organizations unless explicitly authorized, and customer data is never used to train third-party AI foundation models.

Yes. Our frontend applications are ADA compliant, adhering to WCAG accessibility standards, with responsive design that works across desktops, tablets, and smartphones. Contact our support team if you require a specific Accessibility Conformance Report.

Each product has a dedicated dashboard with rich mobility insights: total visitor counts, visitation time series, repeat visitor estimates, dwell times, peak hours, weekday vs weekend and seasonal trends, visitation heatmaps for sub-zones, visitor origins at census block through national levels, routes, transport modes, trip distances and times, and weather correlations.

Yes. Our dashboards support robust side-by-side comparison of multiple parks, trails, facilities, businesses, commercial districts, event venues, road segments, or intersections, so you can see how assets perform relative to one another and how they share visitors.

Yes. You can choose date ranges, toggle metric views and granularities, filter by visitor type such as residents versus tourists, and configure layouts to prioritize the data most relevant to your department. We also welcome feature proposals and can build fully custom dashboards through professional services.

Yes. We offer unlimited on-demand PDF reports, unlimited CSV downloads, spatial exports in KML/KMZ, Shapefile, and GeoJSON formats, direct SQL access to our cloud data warehouse via API, and export pipelines compatible with Excel, Tableau, Power BI, Esri ArcGIS, QGIS, and SAS.

Yes. We provide data exports in standard formats (CSV, JSON, Excel), API integration for automated ingestion into your business intelligence systems, and direct access to cloud data warehouse tables for running SQL queries against our datasets.

Yes. We integrate easily with Esri ArcGIS and other standard GIS platforms, including open-source tools like QGIS. Our geospatial exports in CSV, GeoJSON, and KML formats can be dropped directly into ArcGIS Pro or ArcGIS Online as layers.

Yes. Our exports are fully compatible with Microsoft Power BI and Tableau. You can connect directly to our CSV exports or API to build custom dashboards that combine our mobility data with your other internal datasets.

Yes. While standard dashboards are secured for internal staff, we can configure public-facing 'transparency portals': simplified, read-only views that share key metrics with the community while protecting sensitive internal data. Dashboards can also be embedded in your website and branded with your logo.

Yes. We support enterprise SSO and single log-out with identity providers including Microsoft Azure AD (Entra ID) and Google Workspace, using standard protocols such as SAML 2.0, OAuth 2.0, and OpenID Connect.

As a cloud-native platform, we require no software installation. For transit-specific analytics, we ask for your GTFS or real-time GTFS feeds and GIS shapefiles such as road networks or custom zones. Any voluntary ground-truth data you share, such as sensor counts or ticket sales, further improves accuracy.

Our platforms support hourly, day-part (morning/afternoon/evening/night), daily, weekly, monthly, and annual granularities, enabling analysis from peak-hour congestion down to multi-year longitudinal trends.

CityFlow provides nationwide coverage across all FHWA functional classifications, including interstates and freeways, US and state routes, arterials and collectors, and local roads. It is built natively on OpenStreetMap way IDs, offering segment-level metrics with full coverage of rural and local roads often missing from traditional traffic datasets, and outputs remain compatible with your GIS shapefiles.

Yes. A complete data dictionary with attribute definitions for datasets such as origin-destination movement, trip route trajectories, traffic volumes, and place visits is available at citydata.ai/dictionary.

Yes. All communications use TLS 1.2+ (HTTPS) encryption in transit, and all customer data is encrypted at rest by default using AES-256. Unique identifiers in our mobility data are cryptographically hashed with SHA-256 upon ingestion, so original identifiers can never be reconstructed.

Our platform is hosted on enterprise cloud infrastructure that undergoes rigorous third-party audits and is certified against SOC 2 Type II, ISO 27001, FedRAMP, and PCI-DSS. We enforce least-privilege access controls, conduct annual internal risk assessments and quarterly vulnerability scans, and engage independent third-party firms for annual penetration testing.

For US customers, all cloud resources are configured to reside exclusively in US-based regions, so your data never leaves the United States. We also strictly prohibit storage of customer data on local laptops, USB drives, or mobile devices.

No. CityData has not experienced a single security breach or data compromise since inception. We maintain a formal Incident Response Plan with 24-hour breach notification to affected customers and carry cyber liability insurance coverage.

We maintain a 99% uptime SLA with a Recovery Point Objective of 1 hour and a Recovery Time Objective of 4 hours, backed by quarterly disaster recovery drills, geo-redundant encrypted backups, and long-term archival storage.

Routine updates use rolling deployments with zero downtime, and routine maintenance runs during non-business hours. For major upgrades requiring a maintenance window, we provide at least 72 hours' advance notice; if an unexpected interruption occurs, we notify your technical contacts within 30 minutes and provide hourly updates and a formal root cause analysis.

We follow a rigorous data destruction process aligned with NIST 800-88 standards, including logical deletion from active databases within 30 days of request, cryptographic erasure of encryption keys, and physical destruction of retired storage media by our cloud provider.

Yes. All employees and long-term contractors undergo comprehensive background checks as a condition of employment and complete mandatory security awareness and data privacy training at onboarding, with annual refreshers. Company devices are centrally managed with full-disk encryption and endpoint protection.

No. Our strategic partner is Google Cloud Platform, which provides hosting infrastructure and data center services. We do not use any other third-party subcontractors for application development, data processing, or system operations; all application logic and support are handled by CityData's direct employees.

Yes. Our platform captures granular telemetry including user logins, failed authentication attempts, API calls, and administrative actions. Automated log exports to your secure storage or ad-hoc reports in CSV/JSON format can be configured.

We offer flexible pricing designed to fit government budget cycles, typically structured on the number of assets (parks/places) monitored and population size, with options for annual subscriptions or project-based pay-as-you-go fees. CityFlow uses a fixed-fee subscription model for predictable budgeting, with variables such as geographic extent and optional professional services.

Yes. We highly recommend and support annual billing to simplify procurement for civic clients, though monthly payment terms can be arranged if your agency requires them.

Yes. If you do not require an ongoing subscription, you can purchase a comprehensive historical report for a specific time range to support a master plan or grant application. Bundling historical data with a subscription often results in the most cost-effective rate.

Standard maintenance, updates, and 24/7/365 technical support are included in the subscription fee. Additional fees apply only for work beyond standard scope, such as custom analytics tools or bespoke dashboards, for which we provide written estimates before starting.

We accept major credit cards, bank wire transfers, ACH payments, and traditional mailed checks, and we are experienced with government purchase order processes, providing all necessary vendor forms and W-9s.

We view every contract as a partnership. You will be assigned a dedicated Customer Success Manager who guides you through onboarding, assists with data interpretation, holds regular check-in meetings, and ensures you get maximum value from the platform.

Our core team is based in San Francisco (Pacific Time), with support staff and account managers distributed across multiple time zones to ensure timely responses for customers in North America, Europe, and Asia.

Yes. Standard onboarding includes live virtual training sessions for your team, and we maintain a library of video tutorials, step-by-step guides, and playbooks designed for departments such as parks and recreation. For larger deployments, we can arrange on-site workshops.

We use a professional project management system to log and track every customer request, from feature requests to shape modifications and data inquiries, ensuring accountability and timely resolution.

Our support team is available via email at support@citydata.ai and through in-app chat, and can quickly assist with password resets, account access, and browser compatibility issues.

Yes. We provide a comprehensive User Manual and Installation Guide covering physical mounting through dashboard configuration, so your team can operate the hardware independently.