This article is based on an interview originally published by YourStory, conducted with Gabe Hollombe, Senior Technical Evangelist at AWS, at the India Amazon Web Services Partner Network Award 2019.
Who Is Gabe Hollombe?
Gabe Hollombe is a multilingual full-stack development leader, engineering manager, and software mentor who served as Senior Technical Evangelist at Amazon Web Services (AWS). Having moved from hands-on software development into developer education, his focus is helping software engineers understand how to build on AWS’s ever-expanding suite of cloud services.
“AWS has over 165 services, and the numbers continue to increase each day. I inspire software developers with the possibilities to build with AWS’ ever-growing suite of services,” Hollombe says.
YourStory spoke with Hollombe at the India AWS Partner Network Award 2019 about machine learning (ML), AWS AppSync, Amazon SageMaker, blockchain, and the growing role of IoT in enterprise workloads.
AWS AppSync: Serverless GraphQL APIs in the Cloud
When asked about his favourite AWS technology, Hollombe named AWS AppSync — a managed service that simplifies building GraphQL APIs for storing and retrieving data.
GraphQL is a query language and runtime for APIs that defines how clients should request data from a server, independent of the underlying technology stack. Instead of multiple REST endpoints, a GraphQL API exposes a single endpoint where clients specify exactly the data they need.
AppSync manages the GraphQL endpoint in the cloud so that developers do not have to worry about scalability, security, or high availability — AWS handles those concerns. As Hollombe puts it:
“Developers can now focus more on the ‘what’ and less on the ‘how.'”
Amazon SageMaker: Democratizing Machine Learning
Hollombe’s second favourite service is Amazon SageMaker, AWS’s fully managed platform for building, training, and deploying machine learning models. He frames the core mission clearly:
“Our goal at Amazon and AWS is to put the power of ML in the hands of early developers.”
Before SageMaker existed, many AWS customers wanted to build their own ML models but lacked data scientists or specialists who understood which algorithms to apply. SageMaker addresses this by providing:
- Built-in algorithms covering common problem types such as classification, regression, and clustering, so users do not need to implement algorithms from scratch.
- Managed training infrastructure that scales automatically, removing the need to provision and manage compute resources manually.
- Managed model hosting that serves predictions in a highly available, secure, and scalable environment.
- Documentation and guided workflows that help users select the right algorithm based on the type of problem they are solving.
The result is that developers and analysts without a PhD in data science or advanced mathematics can still build and deploy meaningful ML models.
Real-World SageMaker Example: Intuit
Hollombe cited Intuit as a notable SageMaker customer. Intuit trained a model using SageMaker to help small business owners in the United States automatically itemize taxable deductions from their expenses — a task that previously required significant manual effort or accounting expertise.
High-Level AI Services: Results Without the Complexity
Beyond SageMaker, AWS offers a tier of pre-built AI services that require no ML expertise. Developers supply data and receive results directly. Hollombe highlighted two examples:
- Amazon Rekognition — a computer vision service used for face detection, image moderation, and content analysis. Companies use it to automatically moderate user-uploaded profile images without requiring human reviewers.
- Amazon Polly — a text-to-speech service. Hollombe noted that PolicyBazaar, an Indian insurance comparison platform, used Polly to make automated phone calls to customers with document-upload reminders. The company reportedly observed higher engagement rates from Polly-generated calls than from calls made by human agents.
Blockchain: Separating the Signal from the Hype
Hollombe offered a practical perspective on blockchain, noting that the Bitcoin price spike of late 2017 turned both “bitcoin” and “blockchain” into buzzwords and drove many customers to ask AWS how blockchain could add value to their businesses.
His view is that the right solution depends on the trust model of the use case. AWS offers two distinct services to cover different scenarios:
Amazon QLDB (Quantum Ledger Database)
QLDB is a fully managed ledger database designed for use cases where a single, trusted organization needs a cryptographically verifiable, append-only record of data changes. Every record is cryptographically linked to the previous one, making it mathematically verifiable that historical data has not been altered.
Importantly, QLDB is not decentralized — it is appropriate when all parties trust a central authority. Data can only be appended; past records cannot be modified.
Amazon Managed Blockchain
For scenarios involving a distributed, decentralized network with no single point of trust, AWS offers Amazon Managed Blockchain. This service allows organizations to set up and manage their own private blockchain network without the operational overhead of configuring and maintaining the underlying infrastructure.
Hollombe also mentioned support for Hyperledger Fabric and noted that AWS Ethereum support was forthcoming at the time of the interview (mid-2019).
Enterprise Trends: ML Meets IoT
Hollombe observed that SageMaker adoption was not limited to startups — large enterprises were increasingly using it as well. He also pointed to the convergence of machine learning and IoT as a significant trend.
As more devices — not just smartphones and computers, but industrial sensors — come online, the volume of machine-generated data grows rapidly. Companies are applying ML to this data to enable use cases such as predictive maintenance.
As an example, Hollombe described the Rotimatic, an automated roti-making device whose manufacturer runs on AWS. Each time the appliance operates, it sends device telemetry to the cloud. This data can be analyzed to predict when components are likely to fail, allowing for proactive maintenance before a breakdown occurs.
“Having more data means leveraging it for newer and better customer experience.”
— Gabe Hollombe, Senior Technical Evangelist, AWS
Key Takeaways
- SageMaker is AWS’s primary tool for democratizing ML, offering built-in algorithms and managed infrastructure so that developers without deep ML expertise can still build and deploy models.
- AppSync makes it straightforward to run managed, serverless GraphQL APIs on AWS, letting developers focus on application logic rather than infrastructure.
- High-level AI services like Rekognition and Polly provide ready-to-use AI capabilities via simple API calls.
- AWS’s blockchain strategy distinguishes between centralized, trust-based ledgers (QLDB) and decentralized, multi-party networks (Managed Blockchain).
- The convergence of IoT and ML is creating new enterprise use cases, particularly around predictive maintenance and operational efficiency.