By leveraging the latest technologies in Machine and Deep Learning and state-of-the-art algorithms like Transformers, we're able to improve search results, analyze what your customers are saying about your organization on social media, automatically classify documents, structure information from natural text and more, all tailored to your needs.
Using natural text as input we extract and structure valuable and actionable information. We can help understand unstructured documents, structure information and more.
Social media analysis
Automatic text classification
Form and resume automatic parsing
Automatic captioning and translation
Question and answering
PedidosYa is part of Delivery Hero and they are the market leader for food delivery in LATAM. They are located in multiple countries in the region and are expanding by buying competitors like Glovo.
The client has a vast and diverse food catalog of all restaurants in the region that offer their services through PedidosYa application.
We were given the task to improve existing categorization and to extract further structured information that could allow PedidosYa improve their search results, recommendations and decisions.
It was a typical natural language processing problem.
In this case data was vast but it was not labeled. Labelling the entire dataset was not a possibility because of time constraints. Natural language complexities were abundant as data was inputted by small restaurant owners following, and language variations from country to country only made things worse.
What was a sandwich in some locations was an emparedado in others and when some ice cream shops sell by kilogram, others sell by litre. Although french fries are usually a side dish, if sold alone they can be a plate you share with others and what is called Peruvian cuisine in Argentina is just a typical plate in Peru.
Automatic email classification
We worked with a logistics company that receives a large volume of inquiries via email. Queries must first be classified and registered so that our client's agents can process them. Our client's staff spent hours manually sorting these emails. In addition, this task must be carried out in real time in order to comply with the SLAs for handling these orders.
To automate this time-consuming task, we designed and implemented a classification and registration system to hanflr orders and complaints from our client's account executives. By using a natural language processing engine, we were able to interpret the reason for the request and thus be able to classify it automatically.
Technologies leveraged include Golang, React, AWS, Gsuite integration
Some of the architectures we've worked with
Transformers & BERT
We can fine tune state-of-the art natural language models using Google BERT or even train and develop a Transformer model tailored to your needs.
GPT-2 & GPT-3
We have access and experience working with novel natural language models from OpenAI: GPT-2 and GPT-3.
Use Pegasus, a state-of-the-art model for text summarization from Google AI.
Technologies we use
Get in touch with one of our specialists.
Let's discover how we can help you
Training, developing and delivering machine learning models into production