Highlights
- Advanced Solution for Transportation Construction: We developed an all-in-one platform for HaulHub that incorporates and optimizes the entire transportation construction ecosystem. The solution unifies the processes of agencies, contractors, vendors, and suppliers, ensuring smooth collaboration and alignment among all stakeholders.
- AI/ML Development: Our team implemented advanced AI and ML models to enhance several functionalities across the platform, including NLP for chatbots and image processing for traffic analysis.
- Data Engineering: We overhauled the existing data infrastructure to handle large datasets, which also allowed for real-time processing and visualization. Our data engineering efforts focused on the development of scalable data pipelines, the implementation of a robust Data Lake, and the integration of third-party systems.
- Ongoing Collaboration with HaulHub: Our long-lasting cooperation with HaulHub helped refine and expand the platform’s capabilities to meet evolving transportation construction industry standards and client needs.
Client
HaulHub is a B2B2C company providing a digital platform for the transportation construction industry. The organization is a leader in research and development in their niche. It focuses on pioneering data collection and automation technologies for government and industrial sectors. HaulHub has combined its four advanced solutions into a single multifunctional platform that is designed specifically for heavy construction businesses. In such a way, the company transformed management and digital ticketing processes for material producers, general contractors, suppliers, and carriers across the United States.
Known as a pioneer in the field, HaulHub has partnered with more than 500 contractors nationwide as well as more than 35 state agencies. Many transportation companies leverage the platform as well, using it as a key management tool for digital ticketing operations. HaulHub is acknowledged for its efforts to transform the construction material supply chain and promote accuracy, efficiency, and innovation in the industry.
Country 
Industry 
Team Size: Product
EDOT is a unified portal that optimizes processes within transportation construction projects. This system, created by HaulHub, automates business operations for key stakeholders in construction projects. The platform components seamlessly connect these stakeholders and make it easier for all parties engaged in construction to collaborate. It also facilitates such tasks as digital inspections of operations taking place on site, departmental coordination, and real-time progress tracking.
The platform addresses the needs of four different groups of stakeholders, and for this reason, is built around four interconnected components:
- Agencies (DOT Slip): A hub that streamlines collaboration and digital inspections, automates work zone management, and enhances coordination across departments.
- Contractors: A collaboration tool that enables contractors to manage work zones efficiently, share project data, and collaborate effectively with public owner partners.
- Vendors: A system designed to integrate vendors into the workflow of construction agencies, allowing them to upload submittals, navigate joint projects, and maintain detailed audit logs.
- Suppliers (SupplierCI): It is an analytical system that tracks a wide range of activities from plant to job site, providing near real-time data and insights without relying on GPS.
Goals and objectives
HaulHub was looking for assistance in expanding their mission-critical software platform. We have augmented their in-house team to support the development of all the platform’s components.
Initially engaged by HaulHub, our team discovered an existing component, DOT Slip, designed for agencies. However, the core objective was to develop a solution for the broader ecosystem encompassing construction agencies, contractors, vendors, and suppliers. As the project progressed, we encountered a series of complex feature requirements demanding not only robust software development but also advanced applications of AI/ML and data engineering expertise.
So, we were required to meet the following goals:
- Overhaul of DOT Slip: Improve a management system for transportation agencies by fine-tuning functionalities such as project tracking, digital inspection, and collaboration tools as well as connect the DOT Slip to the ecosystem of contractors, inspectors and suppliers.
- Automate Reporting: Use AI to analyze daily work reports and produce custom overviews aligned with client objectives.
- Enhance Data Presentation: Incorporate widgets customized for both mobile and web platforms.
- Improve Customer Visibility through Data: Provide accurate data on tonnages, materials, projects, plants, trucks, and more to offer a complete view of construction processes.
- Deliver Custom Dashboard and Watchlists: Facilitate decision-making through custom dashboards and watchlists that allow users to monitor top-performing plants and customers.
- Simplify Material Ticketing Data Visualization: Present material ticketing data in a more concise way to streamline its visualization and provide insights into how to optimize resources and ensure additional margin.
- Fine-Grained Project Workload Insights: Provide detailed information on project workloads to allow for precise truck tracking and project completion time estimation.
- Image Processing and Traffic Analysis: Utilize AI to examine construction images in order to guarantee worker safety and pinpoint places in need of process enhancement. Use AI to examine traffic patterns in various construction zones to get an understanding of how road traffic is affected by construction operations before, during, and after construction.
- Develop Chatbots: Use AI to enable communication between clients and the data layer. Make AI choose what information to retrieve from the HaulHub database, get it ready for the client, and respond appropriately.
Project challenge
As we began working on the project, we were asked to develop a set of new functionalities. These new features had to seamlessly integrate with the existing ones. However, the existing functionality relied on legacy code. Therefore, we were required to rewrite most parts of outdated code because of its non-compliance with industry standards and inability to cater to the necessary functionality. Moreover, a larger portion of the work was ahead, including:
- Integration of Complex Business Logic: HaulHub had a complex project, and it was necessary to integrate intricate business logic to address the nuanced requirements of the project and enhance the platform with third-party services. We needed to focus on accommodating complicated data aggregation use cases that required the creation of a flexible system architecture.
- Vast Ticketing Data Volume: The platform needed to manage over 70 million ticketing records. This task included handling the steady increase in dataset volume, processing new tickets continuously in near real-time, and guaranteeing timely updates.
- Client-Facing OLAP System Development: Another responsibility was to build an OLAP system that would be client-facing and optimized for near real-time data availability and performance. This required combining ticketing data from the past and present throughout a number of dimensions and time periods.
- Seamless Authorization: Users lacked the necessary authorization. We configured separate accesses for agencies, vendors, suppliers, and contractors.
- Development and Training of AI Models: To ensure the traffic analysis functionality and guarantee people safety on construction sites, we needed to work on training of several AI models. This was a challenging task because of the long process of understanding the business requirements nuances and converting them into a format that AI could understand.
- Data Analysis: Large datasets needed to be quickly analyzed by AI in order to extract the key information that could be utilized to enhance the construction process.
- Scalability: There were a ton of projects, eTickets, and images stored in the platform. In order to guarantee system stability with a substantial volume of data, we had to carry out intricate tasks.
Solution
To address HaulHub’s need for the whole transportation construction ecosystem, we created mobile and web applications for the EDOT platform. This solution encompasses all four components required for all stakeholders, agencies, contractors, vendors, and suppliers.
We developed the web platform with Java using Quarkus in AWS Claude Lambda. To achieve optimal efficiency we complemented Quarkus by GraalVM for native images. Mobile applications were created using native technologies for iOS (Swift) and Android (Kotlin), since this type of development is a way to ensure the best performance. In its turn, the API layer was developed using a serverless architecture for scalability. Our software engineers also configured monitoring systems to proactively address potential issues.
Data Engineering
The overhaul of the DOT Slip component began with a thorough analysis of the existing code and functional requirements. After improving code quality, we integrated data from internal and third-party systems, ensuring seamless data flow and visualization. Our data engineers built scalable data pipelines and optimized the platform’s data architecture to handle high volumes of projects and eTickets. We also created custom data visualization tools using React.js, Chart.js, Recharts, and Mapbox, providing users with clear and actionable insights.
For the SupplierCI component development, we established a high loaded PostgreSQL, optimizing database performance with techniques such as data denormalization and efficient caching. Data integration was managed through AWS Database Migration Service (DMS) and custom ETL processes, transforming raw data into structured, relational formats for rapid access.
AI/ML Development
We used AWS Lambda as the basis for our client-facing APIs development. These APIs enabled clients to make requests, which were then processed by AI services to produce results based on the provided data.
We employed a range of AI models to enhance the platform’s capabilities. Particularly, we used:
- OpenAI ChatGPT for NLP tasks such as user queries response generation.
- AWS Bedrock Claude/Haiku for summarization, document comparison, and drafting reports.
- AWS Bedrock Titan for content generation, image processing, and search and recommendation systems.
- AWS Bedrock Jurassic for question answering, text generation, and summarization.
We also implemented a Retrieval-Augmented Generation (RAG) approach to improve the relevance and fluency of chatbot interactions.
-
AWS Lambda -
AWS RDS -
PostgreSQL -
AWS S3 -
AWS Batch -
ElastiCache -
Terraform -
GitHub -
CircleCi
-
React.js -
Material UI -
Ruby on Rails
-
Swift -
Kotlin
-
Java -
Quarkus -
GraalVM -
MyBatis -
OpenAI
-
Sentry -
Okta -
DataDog
-
Chart.js -
Recharts -
Mapbox
Our results
We successfully developed web and mobile applications that empower workforce in the transportation construction industry with enhanced operational processes. The platform is highly efficient, handles large sets of data, and is capable of processing millions of data points in under two seconds.
The modern, comprehensive UI/UX we developed offers users the flexibility to create tailored user journeys, while the integration of AI models has enhanced the platform’s functionality, from digital inspections to automated reporting.
Particularly, our efforts resulted in a top-notch platform because we:
- Delivered Key Features: We leveraged our software and data engineering expertise to enable such platform capabilities as digital inspection, live construction activity view, activation and deactivation of digital work zones, worker presence analysis, automated traveler alerts, environmental product declarations reporting, as-built tracking, and more.
- Integrated AI Models: The system is powered by AI and ML models, specifically OpenAI ChatGPT, AWS Bedrock Claude, AWS Bedrock Titan, and AWS Bedrock Jurassic. These models serve purposes such as image processing, report generation, data extraction and classification, traffic analysis, and chatbot functionalities.
- Ensured Stable Integration with External APIs: We designed the platform to function within the broader ecosystem of transportation management tools thanks to enhancements by third-party services.
- Secured the EDOT Portal with Correct Permissions and Roles: Users can login to one portal and get access to all main HaulHub services that include agencies (DOT Slip), contractors, vendors, and suppliers (Supplier CI).
Highlights Client Our client is one of the premier U.S. asset management companies that employs more than 12,000 people,...
Explore Case