How You Can Harness Emerging Tech on a Budget

An interview with Ben Butler, Principal Business Development for Cloud Innovation Centers & Accelerators, AWS Worldwide Public Sector

There’s a growing buzz in government around potential use cases for emerging technologies such as artificial intelligence (AI) and the Internet of Things (IoT). Some agencies have already launched successful pilots and are looking to move their projects into the production phase.

But ramping up these capabilities requires both the technical expertise and infrastructure to support computing, data storage and other needs. To manage these demands, many agencies are forging relationships with third-party vendors to complement their in-house expertise and provide the necessary IT resources and capabilities.

GovLoop sat down with Ben Butler, Principal Business Development for Cloud Innovation Centers & Accelerators for Amazon Web Services (AWS) Worldwide Public Sector, to learn how AWS is enabling emerging technology adoption across government.

“AWS’s machine learning services and platforms allow customers of all sizes to create innovative artificial intelligence applications, and it is all available on-demand at low cost,” Butler said. “AWS cloud computing merges the compute, storage, networking, and database with expertise in specific technologies to provide a robust capability that is simple for our customers to use.”

At a time when agencies are operating with limited budgets and resources, AWS enables them to take advantage of secure, deep learning capabilities without having to start from square one.

For example, agencies can leverage Amazon Rekognition for facial recognition, sentiment analysis, and object detection without having to know the computer vision models or do data preparation that’s required to improve the algorithms. The cable and satellite television network C-SPAN uses the solution to identify who is speaking at on-camera congressional hearings and at what time — down to the second. By using AWS, C-SPAN was able to double its capacity to easily index an archive of more than 97,000 congressional hearing videos.

At the State Department, officials there are planning to deploy a machine learning-based platform and software application suite on AWS GovCloud (US). This isolated region is designed to host sensitive data and regulated workloads in the cloud. Once the new capabilities are in place, the department will be able to predict impending failures of critical facility equipment; automatically monitor, analyze, and manage energy usage across all assets; and assess the health of the sensor and device infrastructure.

“Startups can also leverage AWS and gain access to technology on-demand, at a low cost,” Butler said. “This enables entrepreneurs and employees in government to try out ideas without having to break the bank. Low costs to experimentation increases the frequency of those experiments, which then increases the chances for significant innovation.”

For example, within the United States Institute of Peace in Washington D.C., there is a PeaceTech Accelerator sponsored by AWS, C5 Accelerate, PeaceTech Lab, and SAP NS2. The program provides a space for startups to connect with mentors to reach the next stage of development. These startups are working on high-impact projects, such as using blockchain technology to make transactions transparent and reduce corruption.

In terms of IT infrastructure, another benefit that agencies are taking advantage of is serverless computing. Serverless enables agencies to focus on their business logic because the vendor oversees the provisioning and management of operating systems, applications and servers.

Agencies can create microservices with AWS serverless offerings, such as computing, streaming data in real time, using NoSQL databases, and accessing durable storage that can scale up or down in a matter of milliseconds.

NASA Johnson Space Center is among the agencies taking advantage of a serverless architecture. More specifically, the space center is using a serverless architecture solution to perform optical character recognition and indexing of more than 100,000 pages of spacesuit safety and test documentation each month. The NASA solution enables users to put documents into Amazon S3, which is an object storage built to store and retrieve any amount of data from anywhere. Once the documents are uploaded, a swarm of AWS Lambda functions split the documents and run optical character recognition software on each page in parallel and put the text files into AWS ElasticSearch for quick reference.

“There’s definitely a network effect of using these technologies,” Butler said. “As agencies familiarize themselves with all the capabilities, they can launch new use cases that support their missions.”