Kavak Uses AWS to Power Innovative Machine Learning ApplicationsBy Alejandro Enríquez | Thu, 04/29/2021 - 15:13
Amazon Web Services (AWS) and Kavak, the first Mexican unicorn, implement 'High Gear' Machine Learning (ML) and a serverless architecture. "It is astonishing how much you can accomplish and scale if you truly understand how machine learning can enhance; in a lot of cases, it can completely take over the decision-making process," writes Carlos García, CEO and Co-founder of KAVAK.
AWS explains that the consumer auto industry "has been slow to find its footing in the digital age. It almost sounds too good that a car could show up at your door with your name on it, unless there was a way to bypass the rigmarole of the buying experience. Enter Kavak, the tech startup that’s in the fast lane toward the future of automobile sales," writes Mike Tom of AWS.
Kavak can recommend models based on the client’s needs and guarantee that everything is in shape. The company now offers credit and financing solutions through its new program Kavak Capital. Data science has helped the company to escalate its business to 17 locations in Mexico and more in Argentina after purchasing checkars. Machine learning has been key to optimizing processes and "leveraging a serverless-first infrastructure."
“We use machine learning in many parts of our business including pricing, recommendations, credit scoring and process automation. For example, one model estimates the probability that each auto part will need to be replaced, taking into account the car’s age, mileage, brand and other factors,” says Anders Christiansen, VP of Data Science at Kavak. "It is undeniable that machine learning and AI in the next years will exponentially multiply what disrupters can accomplish through internet and software," agrees García.
The Role of AWS
AWS has played an important role in the development of this solution. It is not the first time the logistics giant has embarked with the auto sector. It has partnerships with Volkswagen AG and other major Tier 1 companies. In the case of Kavak, AWS Glue, AWS Lambda, AWS CloudFormation, Amazon Kinesis Firehose and Amazon CloudWatch allowed Kavak's engineers to put their own models into production to save time and manpower. The process is described in the following image.
According to Chritiansen, Kavak’s next steps are computer vision applications. Today, the company has automated some manual processes but computer vision can be enhanced through more granular information from vehicles' cameras.
For García, machine learning will continue to be an absolute priority for Kavak's growth. "Machine learning is helping us unlock our mission to democratize car ownership for everybody, since day one we took AI seriously, but now we will multiply our efforts and budgets 10 times to make sure we can accelerate the future we are building," he wrote on LinkedIn.