BACK TO CASE STUDIES 
Automating transport order acquisition with AI (80% of orders)
What was the project about?
The client learned about Altkom AI Assist, our virtual assistant solution for incoming and outgoing mail, and contacted us with a project query. They wanted to automate their personalised e-mailing campaigns to reach out to customers who have not contacted the company in a while.
The client wanted to make sure no human intervention would be needed to handle the correspondence and hammer out potential shipment details. All information was to be passed on to a separate application, where sales reps could consult conversation histories, continue processing placed orders or intervene in matters where human involvement was necessary (e.g. when a customer requested a phone call).
march – july 2024
TSL
- AI
- Process automation
- Software development
We can help also your company
Who was our client?
Our client is a modern TSL company specialising in domestic and international full truckload road transport (up to 24 tonnes) and general cargo transport (vehicles of up to 3.5 tonnes). It has offices in several Polish cities and provides services all over Europe.
What were the key challenges of the project?
Altkom AI Assist is a solution based on generative AI, equipped with a language model that is:
- Adapted to the specific demands of the industry, as well as the individual requirements and needs of the client (e.g. the type of product);
- Trained to sound natural in communication with the customer, with a focus on understanding questions and generating a variety of responses;
- Developed to accurately interpret the intention behind each conversation and assess its potential within the framework of every project.
How did we address the problem?
When we joined the project, we already had an AI language model able to communicate naturally in several different languages, understand intentions and react to a lot of situations it had been trained to handle. It only needed to be fed with our client’s data so as to better understand the conversational context and carry out mailing campaigns on behalf of our client’s employees.
For the purposes of the project, we developed a light web app similar to a CRM-class system, to enable company employees to quickly find their way around the new solution. The goal of the app was to store all the information relevant to the campaign, including conversation history or conversation status, in one place. The app was equipped with an AI model with predefined transport order parameters.
In this way, AI Assist now spontaneously sends out mailing campaigns to a list of customers provided by the client, asking them whether they require transport services. It continues the conversation until it has collected all the necessary information (loading/unloading dates and sites, transport type, cargo weight and dimensions). It is able to assess and terminate conversations that do not seem likely to lead to an order and mark conversations that require human intervention.
Are you interested in Altkom AI Assist? Learn more about it
What did we do?
Generative AI
Preparing Altkom AI Assist for training. Initial feeding with materials provided by the client
Fine-tuning
Fine-tuning the model with more domain-specific data sets to improve its grasp of the subject area
Web app
Developing a simple app resembling the client’s CRM for AI-assisted campaigns
AI evaluation
Evaluating the responses generated by the AI model in campaign simulations
UX/UI
Developing the visual aspects of the app to ensure it is easy to use.
Deployment
Deploying and testing the solution in the client’s environment
What did the client get?
Thanks to Altkom AI Assist, our client can now handle campaigns and acquire new orders without any time restrictions.
Artificial intelligence takes care of c. 80% of all campaigns and transport order data collection processes.
AI never forgets any contact and never drops a conversation mid-way, which improves overall company image.
What technologies did we use?
AI Assist
AWS Lambda
AWS Step Functions
AWS S3
Amazon CloudWatch
AWS SQS
AWS Pinpoint
AWS VPC
AWS DynamoDB
Product ManagerArtificial intelligence made our project incredibly scalable. This technology allows you to process huge data volumes and ensure customised service at the highest level, regardless of the size of your customer base.
Companies that choose to invest in generative AI today get a powerful tool that not only helps them manage current operations but also gives them an important future competitive advantage.
Want similar results?
