Integrations
The Google products we integrate, and the GCP stack we wire in
We integrate Google Maps Platform end to end and wire in the Google Cloud services that make it hold at India scale: BigQuery, Vertex AI, Looker, Cloud Logging and Pub/Sub. One vendor, one stack, so you handle surge without standing up self-managed Kafka or Redis. On top, our own products consume the newest GMP analytics.
Get a demo →The shape of it
One hub, spokes to Google and to Google Cloud
Google Maps Platform is rarely the whole stack. The map data sits next to a fleet backend, an analytics warehouse, a forecasting model and an event bus. The integration work is wiring those together so they behave as one system, not a pile of point connections that each fail on their own.
That is what this page covers. The Google products we integrate, the Google Cloud services we wire in around them, and the Lepton product layer that runs on top. It is the same surface our capabilities page selects from, viewed as the wiring rather than the API picking, and it stays a one-vendor story so you are not stitching together a separate operational stack to make it scale.
What we integrate from Google
Three layers of the Google Maps Platform surface
- 01
Core families · Maps, Routes, Places, Environment
Maps (JS, Android, iOS, Static, Embed, Street View, 3D Tiles), Routes (Routes API, Roads, Route Optimization, Navigation SDKs), Places (Places New, Geocoding v4, Address Validation, Geolocation, Time Zone, Places Aggregate) and Environment (Air Quality, Pollen, Solar, Weather).
- 02
Mobility stack · Fleet Engine and the SDKs
Fleet Engine as the managed backend for vehicles, trips and ETAs, with the Driver, Consumer and Navigation SDKs, the Fleet Tracking Library, Nearby Drivers and Vehicle Search, and Shipment Tracking.
- 03
Gated analytics · Insights into your warehouse
Places Insights into BigQuery, Imagery Insights on the allowlist via Cloud Storage and BigQuery, and Roads Management Insights via BigQuery and Pub/Sub, resold under Trafficure.
The Google Cloud stack we wire in
One vendor, so surge is handled for you
Around the Maps surface we wire in the Google Cloud services that turn it into a production system. Keeping it on one vendor means you handle surge without standing up and operating self-managed Kafka or Redis alongside it.
- 01
BigQuery · Trip and market analytics
The warehouse where trip data and the gated GMP analytics land, queryable at scale rather than scraped out of an API one page at a time.
- 02
Vertex AI · Demand forecasting
Forecasting on the location and trip data already in your warehouse, so demand signals feed the same stack everything else runs on.
- 03
Looker · Operations dashboards
Operational dashboards on top of the warehouse, for the teams who run dispatch, fleets and markets day to day.
- 04
Cloud Logging · Observability
Logging across the integration, so usage, errors and behaviour are visible in one place rather than guessed at.
- 05
Pub/Sub · The event bus
The managed event bus that moves data between services, the role you would otherwise self-manage with Kafka, kept inside the same stack.
The wiring, end to end
Where each surface lands, and what has to exist first
The Google surfaces and the Google Cloud services on one sheet, arranged the way a platform team has to plan them: what lands where, what reads it once it is there, and what has to be granted before the first call.
| Surface we integrate | Where it lands | What reads it there | In place before the first call | Lepton product on top |
|---|---|---|---|---|
| Core API families Maps, Routes, Places, Environment Read by Your product surfaces, and Cloud Logging for usage and errors Needs A Google Cloud project, API key or service account, billing enabled | Your own application, over REST and the client SDKs | Your product surfaces, and Cloud Logging for usage and errors | A Google Cloud project, API key or service account, billing enabled | none; we deliver the integration and you own what runs on it |
| Mobility stack Fleet Engine, Driver, Consumer and Navigation SDKs Read by Your driver app, your rider app and your operations console Needs Service accounts and JWT authentication, plus a Mobility Plan | Fleet Engine, the managed backend for vehicles, trips and ETAs | Your driver app, your rider app and your operations console | Service accounts and JWT authentication, plus a Mobility Plan | none; we deliver the integration and you own what runs on it |
| Places Insights Aggregated place counts by area Read by SQL, Looker dashboards, Vertex AI models on the same warehouse Needs BigQuery enabled, dataset access granted | A BigQuery dataset in your project | SQL, Looker dashboards, Vertex AI models on the same warehouse | BigQuery enabled, dataset access granted | SmartMarket |
| Imagery Insights Street View imagery, analysed Read by Vertex AI, for asset identification and condition assessment Needs Allowlist approval, a Cloud Storage bucket and a BigQuery dataset | Cloud Storage, with the derived tables in BigQuery | Vertex AI, for asset identification and condition assessment | Allowlist approval, a Cloud Storage bucket and a BigQuery dataset | none; we deliver the integration and you own what runs on it |
| Roads Management Insights Road network and traffic analytics Read by Looker dashboards and road-authority workflows Needs A BigQuery dataset and a Pub/Sub subscription | BigQuery, with Pub/Sub carrying the stream | Looker dashboards and road-authority workflows | A BigQuery dataset and a Pub/Sub subscription | Trafficure |
Built for India
The India-specific capabilities we lead with
Several Google Maps Platform capabilities exist precisely because Indian addresses, roads and last-mile behave differently. These are the ones we integrate first for India teams.
- 01
Addresses · Address Descriptors and Address Validation
India-exclusive landmark hints relative to a pin, plus address verification at entry and house-level pin accuracy updates, so a wrong address is caught before it becomes a return to origin.
- 02
Two-wheeler · Two-wheeler routing
Bike-specific routing with narrow-roads and flyover logic, for the vehicles that actually run India's last mile.
- 03
Routing · Tolls and live-traffic ETA
Toll and live-traffic awareness in Routes, so quoted arrival times match what happens on the road.
- 04
Places · Gemini-powered Places summaries
Generative summaries on Places data, for richer discovery experiences.
- 05
Compliance · MV Aggregator Guidelines 2025 hooks
Fleet Engine feeds State Command and Control Centres, with route-deviation detection and panic-button-ready integration for cab aggregators.
Google and Google Cloud, converging on one hub.
The Maps surface, the Mobility stack and the Google Cloud services wired into a single integration. One vendor, one stack, so surge is handled without a second operational system to run.
Our own layer on Google's data
The Lepton products that run on top
On top of the integration, Lepton builds its own products on Google Maps Platform's analytics surface. These are not APIs we have only read the documentation for; they are surfaces our own products run on in production.
- 01
SmartMarket · Site selection on Places Insights
SmartMarket consumes Places Insights for site selection and trade-area analysis, running on the same BigQuery layer the analytics land in.
- 02
Trafficure · Citywide traffic on Roads Management Insights
Trafficure consumes Roads Management Insights and Google's traffic dataset for citywide congestion insights.
The data moat underneath
Imagery Insights, built on Google's data
Imagery Insights is one of the newest gated analytics products, and according to Google it draws on a Street View dataset of unusual depth. Google describes it as 280B+ Street View images across 100+ countries, paired with BigQuery and Vertex AI for automated asset identification and condition assessment. These numbers are Google's.
Proof
Wired this way, customers ship on it
NoBroker worked with Lepton on its Google Maps Platform onboarding, and reported a 30-40% conversion-rate boost attributed to seamless navigation and enriched property details. The integration behind it is three of the products listed above, not fifty: Places for neighbourhood context, Geocoding for loosely written addresses, and Distance Matrix for commute time on the listing.
- Places
- Geocoding
- Distance Matrix
Tell us your stack. We will wire GMP into it.
Bring your product and your Google Cloud setup. We map the Google Maps Platform surface to it and deliver the integration, from SDKs to BigQuery.
Get a demo →