Dr. Elliot Smith cannot be accused of lacking ambition. A high achiever with a Ph.D. in Electrical Engineering and a specialist in magnetic resonance imaging (MRI) systems, Smith aims to deliver top quality healthcare to anyone in the world — regardless of their location or wealth.

Dr. Smith has already made strides on this journey with his Brisbane, Queensland-headquartered business, Maxwell MRI. “I saw there was a big gap in the market around automating the diagnosis of health conditions,” he says. “Existing processes were typically manual and involved a lot of people.”

Artificial intelligence (AI) and machine learning can remove a key obstacle to scaling out medicine and improve the efficiency and accuracy of diagnosing conditions, the healthcare entrepreneur believes.

“Our grand vision is to build an AI doctor that anyone can receive affordable support from and connect to in order to obtain results,” explains Dr. Smith.

Maxwell MRI presently enables clinicians to submit anonymised MRI scans to a machine learning enabled AI platform to help diagnose prostate cancer. The service is sold to clinicians who can then charge a per-session fee to clients. As well as obtaining results for individual cases, the MRI scans and associated information is used to ‘train’ the platform to deliver accurate diagnoses faster and in a more affordable way than existing systems do.

Dr. Smith and his team started by running a number of functions and processes on a single server with graphics processing units (GPUs) and sizable hard disk capacity. However, this infrastructure could not scale to support the planned growth of the business. Each case Maxwell MRI processes involves about 200MB of data in MRI scans alone. Once supplementary data, blood test result, pathology results and genetic information is included, this load can reach more than 1GB of data per patient.

The business aimed to process 150,000 cases by the end of 2018. This required a service that could deliver massive scale in data storage and compute, and could easily be accessed from any location. “We wanted to move from three GPUs to 30 GPUs without having to buy more servers or other associated equipment, so the cloud was the natural next step,” says Dr. Smith.

Maxwell MRI evaluated Google Cloud Platform (GCP) and determined that the managed services component of GCP would remove the burden of infrastructure deployment and administration. In addition, Google Cloud Machine Learning Engine would enable the business to scale to as many GPUs as needed to meet demand.

Maxwell MRI started with some small experiments to determine that GCP met all its requirements and completed its migration to the platform in February 2017. “We really started to scale up the data we had and consequently our computing requirements at that time,” Dr. Smith says.

The Maxwell MRI platform features an upload service that enables clinicians to upload imaging and associated data. This service triggers several different upload pipelines that clean and standardise data. They then write imaging data to Google Cloud Storage, and more structured data to a combination of Google Cloud Datastore and Google Cloud Spanner.

The platform then converts the information into records that can be used to ‘train’ new machine learning configurations or run evaluations through existing machine learning pipelines.

“The tasks we perform including segmenting various anatomical regions for analysis and sending those results back into Google Cloud Storage,” says Dr. Smith. “This then commences that repeated process of running Google Cloud Dataflow pipelines and machine learning algorithms, and presenting those outcomes back to the clinicians.”

Existing Literature Validated

The data processed and analysed to date has, Dr Smith says, enabled Maxwell MRI to help validate existing literature that indicates clinicians lack confidence in existing early-stage testing procedures for prostate cancer. This prompts them to move quickly to the biopsy stage to assure themselves their diagnosis is valid. “New technologies have a lot of potential to rectify this situation and guide treatment to be more accurate, specific and cost-effective,” he says.

Results Delivered in 10-15 Minutes

More specifically, using GCP has enabled Maxwell MRI to guarantee to clinicians that results will be delivered within minutes. “Clinicians are used to getting results back in two days to a week,” says Dr. Smith. “We’re saying that with our platform running on GCP we’ll deliver you results in 10 to 15 minutes, regardless of the number of patients coming in.”

Running on GCP has enabled the business to accelerate its development cycles, test new ideas easily on a subset of data, test in parallel and deliver new services considerably faster than in another environment. In addition, the flexible GCP charging model aligned with the ability to scale compute capabilities quickly and easily has enabled the fledgling business to control its costs.

Google technologies are poised to play an integral role in the business’s future. “With Google available, it doesn’t make sense for us to use our own infrastructure,” Dr. Smith says. “Our expertise in AI, machine learning and clinical engagement complements cloud platform specialties of infrastructure, managed services and ease of use. We see a bright future ahead in helping to transform healthcare globally.”

As lead for analytics and AI solutions at Google Cloud, my team works with startups building on Google Cloud. This puts us in the fortunate position to learn from founders and engineers about how early-stage startups’ investments can either constrain them or position them for success, even at the seed level. In this post, I want to share a few of the best practices to keep in mind as you’re building. 

Understand your value proposition before diving into a technology stack

If you’re launching a startup in the cloud, you’re no doubt thinking about a technology stack, but it’s important to step back a bit and think carefully about the major value proposition that your startup offers to your customers. That value proposition is going to fundamentally drive the kind of technology that you should pick.

For example, does your system need processing in real time, or can it be done in a batch mode? Can you rely on once-a-day insights or do the insights have to come in as events happen?

Additionally, what kind of latency will your customers face? That latency makes your value proposition either usable or unusable. Early on in Google’s development, leaders realized that no one was going to wait more than a few hundred milliseconds for a web page to show them their results, and that realization drove the technology decisions that have allowed Google to scale from being a startup in a garage to being a trillion dollar company. Your startup needs to define its value to customers with this level of specificity before it can build a technology stack suited to its needs. 

Focus on customer interactions

A few companies have gracefully pulled off big IT pivots that reshaped their value proposition. Netflix, for example, moved from mostly sending DVDs through the mail to becoming a streaming service and major content producer. That’s a huge shift in the user experience and the technology stack necessary to support it, even if the underlying value proposition (i.e., get content to customers) was broadly the same. But it’s also an outlier. If you’re planning for potential changes of this magnitude, rather than focused on getting your value proposition to users, you probably need to sharpen what that value proposition is.

Specifically, you need a clear vision of how customers will access and interact with your business. Typically, they’ll do so over a website or a mobile app, but there are still so many variables. 

Are customers going to transmit documents? If so, in what format? Is handwriting supported or is input limited to typing? Can they use images for optical character recognition? Will it mostly be forms? Will the data be structured or unstructured? If all that sounds  a little overwhelming, don’t worry, it’ll seem simpler by the end of this article—but also be aware: we’re just getting warmed up.

Imagine that most of your customers will access your business via voice, so you know you’ll want to prioritize conversational workflows. That’s a start—but dig deeper.  Even if we suppose you’re usingDialogflow, a Google Cloud conversational AI platform that lets you build and deploy virtual agents, we’re still not really seeing the value proposition.  How will all this work, from the beginning of a typical full customer interaction to the resolution? How many interactions will have to be facilitated over low-bandwidth connections, for example? When it comes to user interactions, make sure you can see an end-to-end use case.

Another example: you’re building a retail website, and one of your end-to-end use cases involves the customer asking if a certain amount of a given product is in stock, whether it’s one unit of the product, ten or hundreds. If the product is not sufficiently stocked, you want your app to offer similar items that are. Will your technology stack support this end-to-end use case?

These considerations are not an argument for premature optimization. There’s value in moving fast, getting minimum viable products to users, and then iterating. But in the early stages, you only get one chance to start on the right foot—and how you navigate that chance will influence a lot of dollars and effort down the road. You need to make sure you have business use cases, not just an idea, before you can start designing a technology stack.  

Here’s how to get in the right frame of mind. Pick three use cases: two that are “bread and butter” and one that is technologically complex.  Make sure your proposed technology stack can support all three, end to end. 

Default toward higher levels of abstraction

Now that we’re in the right frame of mind, we’re ready to think about the technology stack more directly. 

As a startup, you’ll need to conserve resources, and to do that, you’ll want to build at the highest level of abstraction possible for your value proposition. For example, you probably don’t want your people setting up clusters. You don’t want them configuring things if they can use a fully managed service. You want them focused on building your prototype, not managing infrastructure.

This focus has definitely informed how we create products at Google Cloud, as our canonical data stack—Pub/Sub, Dataflow, BigQuery, and Vertex AI—consists of auto-scaling and serverless products.

But management of infrastructure is not the only place where you should err toward a less-is-more philosophy. 

When it comes to architecture, choose no-code over low-code and low-code over writing custom code. For example, rather than writing ETL pipelines to transform the data you need before you land it into BigQuery, you could use pre-built connectors to directly land the raw data into BigQuery. That’s no code right there. Then, transform the data into the form you need using SQL views directly in the data warehouse. This is called ELT, and it is low code. You will be a lot more agile if you choose an ELT approach over an ETL approach. 

Another place is when you choose your ML modeling framework. Don’t start with custom TensorFlow models. Start with AutoML. That’s no-code. You can invoke AutoML directly from BigQuery, avoiding the need to build complex data and ML pipelines. If necessary, move on to pre-built models from TensorFlow Hub, HuggingFace, etc. That’s low-code. Build your own custom ML models only as a last resort.

Focus on getting your vision to market, not chasing technology hype  

The goal is to pick the right technology stack for bringing your vision to market, generating value for customers, conserving resources, and maintaining flexibility for growth. Early IT investments should usually gravitate toward things that preserve flexibility, such as managed services built on standard protocols or open APIs, but they needn’t always rush to the flashiest technologies.  The answer isn’t always ML, for example. The answer might be heuristics to start, with a path to ML once you have collected enough data. You want to make sure that your intelligence layer has enough abstraction so you can mark it up with simple rules at first, but then replace it with a more robust system as you go along. 

Launch and iterate fast with these principles 

The preceding discussion is a reminder that your most expensive resource is your people—and that you really want them to be focused on building your prototype, minimum viable product or production app  You want to launch fast and iterate fast, and the only way you can do that is by focusing on the things that differentiate you. 

But regardless of the technologies you use, the bottom line is the same: follow these four principles. 

• Figure out your major value proposition and design your tech stack around it. 
• Be very careful about user interactions. User experience is super important; you need to make sure you deliver the kind of experience that your customers have grown to expect.
• When you’re building, pick the highest possible level of abstraction possible—the most fully managed tools and no-code/low-code frameworks that give you the functionality that you need. 
• Instead of choosing new or flashy technologies, consider if you can build a “good enough” minimum viable product quickly and come back to a better implementation later. 

To learn more about why startups are choosing Google Cloud, click here.

When solving a new ML problem, it’s common to start by experimenting with a subset of your data in a notebook environment. But if you want to execute a long-running job, add accelerators, or run multiple training trials with different input parameters, you’ll likely find yourself copying code over to a Python file to do the actual computation. That’s why we’re excited to announce the launch of the notebook executor, a new feature of Vertex AI Workbench that allows you to schedule notebooks ad hoc, or on a recurring basis. With the executor, your notebook is run cell by cell on Vertex AI Training. You can seamlessly scale your notebook workflows by configuring different hardware options, passing in parameters you’d like to experiment with, and setting an execution schedule, all via the Console UI or the notebooks API. 

Built to Scale

Imagine you’re tasked with building a new image classifier. You start by loading a portion of the dataset into your notebook environment and running some analysis and experiments on a small machine. After a few trials, your model looks promising, so you want to train on the full image dataset. With the notebook executor, you can easily scale up model training by configuring a cluster with machine types and accelerators, such as NVIDIA GPUs, that are much more powerful than the current instance where your notebook is running.

Your model training gets a huge performance boost from adding a GPU, and you now want to run a few extra experiments with different model architectures from TensorFlow Hub. For example, you can train a new model using feature vectors from various architectures, such as Inception, ResNet, or MobileNet, all pretrained on the ImageNet dataset. Using these feature vectors with the Keras Sequential API is simple; all you need to do is pass the TF Hub URL for the particular model to hub.KerasLayer.

Instead of running these trials one by one in the notebook, or making multiples copies of your notebook (inceptionv3.ipynb, resnet50.ipynb, etc) for each of the different TF Hub URLs, you can experiment with different architectures by using a parameter tag. To use this feature, first select the cell you want to parameterize. Then click on the gear icon in the top right corner of your notebook.

Type “parameters” in the Add Tag box and hit Enter. Later when configuring your execution, you’ll pass in the different values you want to test.

In this example, we create a parameter called feature_extractor_model, and we’ll pass in the name of the TF hub model we want to use when launching the execution. That model name will be substituted into the tf_hub_uri variable, which is then passed to the hub.KerasLayer, as shown in the screenshot above.  

After you’ve discovered the optimal model architecture for your use case, you’ll want to track the performance of your model in production. You can create a notebook that pulls the most recent batch of serving data that you have labels for, gets predictions, and computes the relevant metrics. By scheduling these jobs to execute on a recurring basis, you’ve created a lightweight monitoring system that tracks the quality of your model predictions over time. The executor supports your end-to-end ML workflow, making it easy to scale up or scale out notebook experiments written with Vertex AI Workbench.

Configuring Executions

Executions can be configured through the Cloud Console UI or the Notebooks API.

In your notebook, click on the Executor icon.

In the side panel on the right specify the configuration for your job, such as the machine type and the environment. You can select an existing image, or provide your own custom docker container image.

If you’ve added parameter tags to any of your notebook cells, you can pass in your parameter values to the executor.

Finally, you can choose to run your notebook as a one time execution, or schedule recurring executions.

Then click SUBMIT to launch your job.

In the EXECUTIONS tab, you’ll be able to track the status of your notebook execution.

When your execution completes, you’ll be able to see the output of your notebook by clicking VIEW RESULT.

You can see that an additional cell was added with the comment # Parameters, that overrides the default value for feature_extractor_model, with the value we passed in at execution time. As a result, the feature vectors used for this execution came from a ResNet50 model instead of an Inception model.

What’s Next?

You now know the basics of how to use the notebook executor to train with a more performant hardware profile, test out different parameters, and track model performance over time. If you’d like to try out an end-to-end example, check out this tutorial. It’s time to run some experiments of your own!

Forrester Research has named Google Cloud a Leader in The Forrester Wave™: AI Infrastructure, Q4 2021 report authored by Mike Gualtieri and Tracy Woo. In the report, Forrester evaluated dimensions of AI architecture, training, inference and management against a set of pre-defined criteria. Forrester’s analysis and recognition gives customers the confidence they need to make important platform choices that will have lasting business impact. 

Google received the highest possible score in 16 Forrester Wave evaluation criteria: architecture design, architecture components, training software, training data, training throughput, training latency, inferencing throughput, inferencing latency, management operations, management external, deployment efficiency, execution roadmap, innovation roadmap, partner ecosystem, commercial model, and number of customers.

We believe that Google’s vision to be a unified data and AI solution provider for the end-to-end data science experience is recognized by Forrester, through high scores in the areas of architecture and innovation. We are focused on building the most robust yet cohesive experience to enable our customers to leverage the best of Google every step of the way. Here are four key areas where Google excels, among the many highlighted in this report. 

AI Infrastructure: Leverage the building blocks of innovation

When an organization chooses to run its business on Google Cloud, it benefits from innovative infrastructure available globally. Google offers users a rich set of building blocks such as Deep Learning VMs and containers, the latest GPUs/TPUs and a marketplace of curated ISV offerings to help architect your own custom software stack on VMs and/or Google Kubernetes Engine (GKE). 

Google provides a range of GPU & TPU accelerators for various use cases, including high performance training, low cost inferencing and large-scale accelerated data processing. Google is the only public cloud provider to offer up to 16 NVIDIA A100 GPUs in a single VM, making it possible to train very large AI models on a single node. Users can start with one NVIDIA A100 GPU and scale to 16 GPUs without configuring multiple VMs for single-node ML training.  Google also provides TPU pods for large-scale AI research with PyTorch, TensorFlow, and JAX.  The new fourth generation TPU pods deliver exaflop-scale peak performance with leading results in recent MLPerf benchmarks which included a 480 billion parameter language model.   

Google Kubernetes Engine provides the most advanced Kubernetes services with unique capabilities like Autopilot, highly automated cluster version upgrades, and cluster backup/restore. GKE is a good choice for a scalable multi-node bespoke platform for training, inference and Kubeflow pipelines, given its support for 15,000 nodes per cluster, auto-provisioning,  auto-scaling and various machine types (e.g. CPU, GPU, TPU and on-demand, spot). ML workloads also benefit from GKE’s support for dynamic scheduling, orchestrated maintenance, high availability, job API, customizability, fault tolerance and ML frameworks.  When a company’s footprint grows to a fleet of GKE clusters, its data teams can leverage Anthos Config Management to enforce consistent configurations and security policy compliance. 

Comprehensive MLOps: Build models faster and more easily without skimping on governance 

Google’s fully managed Vertex AI platform provides services for ML lifecycle management, from data ingestion and preparation all the way up to model deployment, monitoring, and management. Vertex AI requires nearly 80% fewer lines of code to train a model versus competitive platforms1, enabling data scientists and ML engineers across all levels of expertise to implement Machine Learning Operations (MLOps) so they can efficiently build and manage ML projects throughout the entire development lifecycle. 

Vertex AI Workbench provides data scientists with a single environment for the entire data-to-ML workflow, enabling data scientists to build and train models 5x faster than traditional notebooks. This is enabled by integrations across data services (like Dataproc, BigQuery, Dataplex, and Looker), which significantly reduce context switching.  Users are also able to access NVIDIA GPUs, modify hardware on the fly, and set up idle shutdown to optimize infrastructure costs. 

Organizations can then build and deploy models built on any framework (including TensorFlow, PyTorch, Scikit learn or XGBoost) with Vertex AI, with built-in tooling to track a model’s performance.  Vertex Training also provides various approaches for developing large models including Reduction Server to optimize bandwidth and latency of multi-node distributed training on NVIDIA GPUs for synchronous data parallel algorithms.  Vertex AI Prediction is serverless, and performs automatic provisioning and deprovisioning of nodes behind the scenes to provide low latency online predictions. It also provides the capability to split traffic between multiple models behind an endpoint. Models trained in Vertex AI can also be exported to be deployed in private or other public clouds.Google’s strengths in its current offering are in architecture, training, data throughput, and latency. Its sweet spot is in its product offering, Vertex AI, which has core AI compute capabilities and MLOps services for end-to-end AI lifecycle management.The Forrester Wave:™ AI Infrastructure, Q4 2021

In addition to building models, it is important to deploy tools for governance, security, and auditability. These tools are crucial for compliance in regulated industries, and they help teams to protect data, understand why given models fail, and determine how models can be improved. 

For orchestration and auditability, Vertex Pipelines and  Vertex ML Metadata tracks the inputs and outputs of an ML pipeline and the lineage of artifacts. Once models are in production, Vertex AI Model Monitoring supports feature skew and drift detection, alerting data scientists. These capabilities speed up debugging and create the visibility required for regulatory compliance and good data hygiene in general.For explainability, Vertex Explainable AI helps teams understand their model’s outputs for classification and regression tasks. Vertex AI tells how much each feature in the data contributed to the predicted result. Data teams can then use this information to verify that the model is behaving as expected, recognize bias in the model, and get ideas for ways to improve the model and training data.

These services together aim to simplify MLOps for data scientists and ML engineers, so that businesses can accelerate time to value for ML initiatives.

Security: Protect data while keeping ML pipelines flowing  

The Google stack builds security through progressive layers that deliver defense in depth. To accomplish data protection, authentication, authorization and non-repudiation, we have measures such as boot-level signature and chain-of-trust validation. 

Ubiquitous data encryption delivers unified control over data at-rest, in-use, and in-transit, with keys that are held by customers themselves. 

We offer options to run in fully encrypted confidential environments utilizing managed Hadoop or Spark with Confidential Dataproc or Confidential VMs.  

Partner Ecosystem: Work with world-class AI specialists 

Google works with certified partners globally to help our customers design, implement and manage complex AI systems. We have a growing list of partners with Machine Learning specializations on Google who have demonstrated customer success across industries, including deep partnerships with the largest Global System Integrators. The Google Cloud  Marketplace also provides a list of technology partners who allow enterprises to deploy machine learning applications on Google’s AI infrastructure.

Our dedication to being your partner of choice for ML Needs  

Leading organizations like OTOY, Allen Institute for AI and DeepMind (an Alphabet subsidiary) choose Google for ML, and enterprises like Twitter, Wayfair and The Home Depot shared more about their partnership with Google in their recent sessions at Google Next 2021.

Establishing well-tuned and appropriately managed ML systems has historically been challenging, even for highly skilled data scientists with sophisticated systems. With the key pillars of Google’s investments above, organizations can build, deploy, and scale ML models faster, with pre-trained and custom tooling, within a unified AI platform. 

We look forward to continuing to innovate and to helping customers on their digital transformation journey. To download the full report, click here. Get started on Vertex AI, learn what’s upcoming with infrastructure for AI and ML at Google here, and talk with our sales team.