Project highlights
- Recognize photos and documents of drivers and transportation providers
- Automate onboarding for accelerated freight management
- Aggregate data on new users to compare against baseline criteria
- Industry:
- Logistics & Transportation
- Expertise:
- AI & ML
- Team size:
- 8 engineers
- Market:
- UK
- Cooperation :
- August 2019 – present
Business challenge
Our client is the provider of a real-time freight marketplace that logistics and transportation companies use to move over 1.3 million shipments around the world each year. This marketplace connects shippers — usually big logistics companies and retailers who need subcontractors for freight transportation — with drivers or smaller transportation companies in search of orders.
The marketplace essentially works like an Uber for trucking application. The app enables couriers to place orders for shipments and allows drivers and small transportation companies to fulfill them.
To assure couriers that their products are in good hands, Uber for trucking companies have to establish baseline criteria for drivers, vehicles, and transportation history. Only after comparing a new driver against baseline criteria, an Uber for trucking app allows the driver to apply. Initially, our client’s support center manually checked all drivers’ and companies’ documents for compliance, which made the onboarding process last for months.
Our client needed to more effectively onboard new businesses and drivers into their freight management marketplace as well as optimize the time and resources spent on the onboarding process. To do that, they searched for a software development partner with expertise in the logistics and transportation industry to cover end-to-end component development and propose an optimal solution for automating the onboarding process by applying advanced technologies.
Solution delivered
Our client entrusted us to develop a complete product ready for delivery to the market that organizes the process of onboarding new businesses into their freight management platform. Together with our client, we organized a series of workshops to decide on the optimal tech stack to cover the project’s requirements. As a result of close collaboration with the product owner and our client’s in-house engineering team, we came up with basic use cases to include in the onboarding process.
The freight management system assigns roles to different users, providing access to role-relevant functionality and organizing optimal onboarding. The basic onboarding scenarios include validating an individual driver with a vehicle and validating a transportation company with a pool of drivers available for delivery orders.
User roles defined for onboarding workflows:
- Driver or vehicle owner who is solely responsible for freight transportation, account management, invoices, and service quality
- Delivery company that covers the transportation process through a corporate account without needing to create individual accounts for each driver
- Dispatcher or fleet manager who represents drivers and administrates their accounts from onboarding to successful delivery of orders
Depending on the user’s role, onboarding may require a different set of documents to open access to relevant functionality. To ensure the most efficient experience for each user of an Uber for trucking app and minimize the involvement of customer support, we focused on automating features to eliminate manual work related to gathering data on new users and validating their documents.
We implemented a new component that automates data aggregation, processing, and decision-making through robust API integrations. This component automatically validates drivers, companies, licenses, vehicle conditions, and insurance plans, eliminating potential fraud.
Users upload documents to our client’s system via an online survey. The system then parses user data to recognize identity of applicant, and the expiry date of the documents. Users also need to take real-time photos that the system then compares against their ID photos using image recognition software based on computer vision and AI that’s specifically trained and optimized for documents and photo processing.
APIs for automated onboarding to an Uber for trucking app:
- DueDil API — Provides access to comprehensive company data with descriptions, general information, and turnover rates for corporate accounts
- IDscan WebSDK — Features rich document scanning, detection of liveness, and face matching functionality to scan the uploaded ID documents of users, parse them, and validate users with photos taken by the API
- GBG Geolocation API — Verifies addresses using the most accurate global location intelligence to autocomplete search inquiries for ZIP code and address
- Vatlayer API — Validates VAT numbers and rates based on IP address or country code and converts prices in compliance with EU VAT rates
- MOT History API — Provides history data on the technical condition of a vehicle using its registration number, including data on mileage, failures, vehicle registration and expiry dates, date of manufacture, and engine size
- Google Geocode API — Provides comprehensive real-time data on points of interest, routes, documentation, and addresses
- HERE Geocoder API — Allows you to get geocoordinates for addresses, administrative areas, and landmarks
- Geocodio API — Finds addresses by latitude and longitude, converts addresses into geo coordinates, and stores customer geodata
After successful onboarding into our client’s Uber for trucking app, users can access a dashboard with all orders. In this dashboard, drivers can choose the most convenient orders to deliver. After a driver chooses an order, the system compares the driver’s profile with baseline criteria. If the user is a match for the job, they’re provided with all the data on the order including the address, customer, product provider, and receiver. Using the Uber for trucking, fleet managers can track orders and send instant messages to drivers to change delivery routes or provide shipment details. Uber for trucking works in mobile applications that help organize a driver’s work on the go, including by synchronizing with navigation systems, while the desktop app is a convenient method for dispatchers to support drivers and manage corporate accounts.
Business outcome
The money and time spent onboarding employees, users, and assets is a burden on big companies. After successfully delivering an automated onboarding system for their freight marketplace, our client is planning to make this component marketable to sell it to other logistics businesses. Our client is paying special attention to data privacy and the security of data collected on drivers’ and companies’ profiles. Security measures they’re taking include mitigating long-term risks, complying with international regulations, and conducting security audits.
To achieve better scalability and efficiency of their solutions, our client plans to extend their engagement with Intellias in an effort to substitute their monolithic platform with a microservices architecture. We’ve already established an initial R&D team to work on technology stacks to migrate our client’s platform to microservices.
Our client’s automated onboarding system has cut the time spent onboarding new users from months to minutes. It eliminates the need to involve the support center to check all documents, saving money for our client, and allows users to start delivering products right away.
Our client continues to grow their business thanks to the introduction of new technologies and expects their revenue increase as a result of these changes to jump from the current 25% to 40% in the near future. While keeping their strategic focus on the UK market, they’re now planning to expand their Uber for trucking app’s reach into continental Europe.