Mantle Labs Improves Global Food Supply Chain Financing on AWS
Satellite Data from 500 Million Hectares of Land on Six Continents
Mantle Labs took advantage of artificial intelligence (AI) to build its flagship Geobotanics platform, which analyzes data feeds from satellites in space. This data is used to develop agricultural risk and decision-making indices specific to industry sectors linked to agriculture: banks, insurers, food retailers, crop input providers, and commodity companies that either trade or manage farms directly. The insurance index, for example, monitors crop growth to compare against historical patterns, also factoring in external data on past and projected weather conditions to shape the index.
By subscribing to Geobotanics, banks and insurers can make more informed decisions on rates for agri-loans and farming insurance premiums. Farmers benefit from new and improved sustainable finance solutions, and commodity companies, food retailers, and other members of the agricultural supply chain can better plan their businesses and prevent food shortages.
Mantle Labs uses Amazon Web Services (AWS) to automate processing of satellite data from 500 million hectares of agriculture spanning six continents. By building Geobotanics on AWS, Mantle Labs has drastically reduced the manual effort required to process high volumes of data using its AI algorithms. The company has access to on-demand infrastructure that can be scaled up and down seamlessly, and it can stay at the cutting edge of technology while controlling costs.
With AWS, we are constantly operating on the edge of what’s possible. We’re encouraged to keep up to date with the latest technology to convert our concepts into real-world solutions.”
Cofounder, Mantle Labs
Cost-Effective, Large-Scale Data Processing
The founders of Mantle Labs had a vision to develop the Geobotanics solution years before they brought the product to market. Satellite imagery data is extensive, and until about eight years ago, the technology required to access such a high volume of data would have been cost-prohibitive. “We realized that large-scale data processing can be handled on AWS in a very cost-effective way, which turned our vision into reality,” says Swapnil Baokar, cofounder of Mantle Labs.
Every week, Mantle Labs processes 32–50 TB of data from satellites around the globe. “The availability of on-demand, large-scale infrastructure was critical for this platform’s success. AWS is able to provide that kind of capacity without going through any provisioning bureaucracy,” adds Baokar.
During peak data processing, Mantle Labs can easily run 1,000 or more virtual machines simultaneously. Accessing satellite data to initiate processing is likewise seamless. Imagery downloaded from the National Aeronautics and Space Administration (NASA) and other space agencies are available to AWS customers via Amazon Simple Storage Service (Amazon S3) buckets.
Automation Saves Time across Data Pipeline
Five years into operations, Mantle Labs had tripled its business. Founders then consulted AWS on how to improve performance to handle the increased capacity in a more cost-efficient way. AWS introduced the company to Comprinno Technologies, an AWS Partner Network Advanced Consulting Partner, to help with optimizing its cloud architecture. Comprinno worked with Mantle Labs to integrate automation into nearly every stage of the data pipeline.
The business containerized its architecture and started using serverless functions to enable infrastructure as code. Before this modernization, Mantle Labs’ data science team had to sort through large image feeds to prepare data for processing specific client requests. For example, the main AI engine they used to fill data gaps for cloud cover took 7–10 days to process and stand up. It takes less than 15 minutes to prepare the Helios AI engine for model deployment.
Empowered with DevOps capabilities, the data science team no longer requires as much IT support to test, set up, and run AI processing models. “With the automation now in place, the data science team can run the infrastructure on their own, which has eliminated a lot of the delays and friction that resulted from manual processes,” says Rishikesh Sapre, cofounder of Mantle Labs. “We’ve seen this beautiful synergy emerge between the data science and infrastructure teams as a result of this data-as-a-service approach.”
Base Cost Reduction Enables New Pricing Models
To address the rising costs of data processing, Mantle Labs started using Amazon Elastic Compute Cloud (Amazon EC2) Spot Instances. It has now reduced costs by 40 percent compared to using Amazon EC2 On-Demand Instances. These savings are passed onto customers in the form of new pricing models. “Initially, we had a base cost for processing, but we’ve achieved optimization at scale to spend less with every action performed on the Geobotanics platform. We’ve started offering new engagement models via user license-based solutions to clients that are smaller in size, whereas previously we could target only large companies,” explains Baokar.
Several farming regions in Africa have also gained access to insurance for the first time thanks to the data-backed indices on Geobotanics. Before, many insurers could not offer coverage because they didn’t have enough information to decide on appropriate premiums. And those that could offer coverage sold it at premiums valued at 12–15 percent of estimated crop production market price. Now, even small-scale farmers can get insurance at rates as low as 3–5 percent of crop market price.
Flexible Deployments with Cutting-Edge Technology
With AWS, Mantle Labs has the flexibility to choose among deployment methods and instance types to suit individual workloads. It uses AWS Fargate as a serverless compute engine and Amazon Elastic Container Registry (Amazon ECR) for storing container images. Mantle Labs also leverages Amazon Simple Queue Service (Amazon SQS) for event-based batch processing of satellite images.
“The complexity of the Geobotanics platform could only be achieved on AWS,” Sapre says. “With AWS, we are constantly operating on the edge of what’s possible. We’re encouraged to keep up to date with the latest technology to convert our concepts into real-world solutions.”
To Learn More
To learn more, visit aws.amazon.com/big-data/datalakes-and-analytics.
About Mantle Labs Limited
Mantle Labs is a remote-sensing company providing digital agriculture solutions backed by artificial intelligence. Its industry-specific risk indices help banks and insurers make critical decisions that impact farmers and other players in the global food supply chain.
Benefits of AWS
- Processes 32–50 TB of data each day
- Reduces cost by 40% for AI-powered data processing
- Can readily access satellite imagery data
- Reduces time required for standing up AI models from 7 days to 15 minutes·
- Empowers its data science team with DevOps tools
- Facilitates access to loans and insurance for small-scale farmers
- Develops a new pricing model to serve clients of all sizes
AWS Services Used
AWS Fargate is a serverless compute engine for containers that works with both Amazon Elastic Container Service (ECS) and Amazon Elastic Kubernetes Service (EKS). Fargate makes it easy for you to focus on building your applications.
Amazon Elastic Container Registry
Amazon Elastic Container Registry (ECR) is a fully managed container registry that makes it easy to store, manage, share, and deploy your container images and artifacts anywhere.
Amazon EC2 Spot Instances
Amazon EC2 Spot Instances let you take advantage of unused EC2 capacity in the AWS cloud. Spot Instances are available at up to a 90% discount compared to On-Demand prices.
Amazon Simple Storage Service
Amazon Simple Storage Service (Amazon S3) is an object storage service that offers industry-leading scalability, data availability, security, and performance.
Companies of all sizes across all industries are transforming their businesses every day using AWS. Contact our experts and start your own AWS Cloud journey today.