With the rise of technologies like smartphones, retailers have felt the pressure to meet evolving consumer needs and expectations. Seven-Eleven Japan(“SEJ”) has long been on the forefront of this thanks to the way they develop and invest in IT. However, in recent years, Japan’s leading convenience store chain has struggled to maintain its complex legacy systems at the rate needed to keep up with today’s rapid digitization, spurred on by the increasing proliferation of smartphones and an IT vendor-dependent structure.  

Legacy systems limiting real-time responsiveness and innovation 

Since its early days, SEJ has been proactive in adopting information technology, mainly relying on technology solutions from Japan’s leading vendors. But as the systems have grown, key business issues have been resolved using a vendor-dependent structure rather than being driven by SEJ’s own needs.  

Datasets and business logic were combined and built into legacy environments, gradually leading to data silos. As a result, data was distributed across multiple systems, causing a variety of problems, including the inability to efficiently retrieve data when needed, delays in accessing data collected in individual stores, and difficulties taking measurements at the right time in business operations that require real-time responsiveness.

Connecting different systems also takes time and money, and the lead time for introducing new services—from planning to development and launch—has been longer than expected. 

To solve these problems, SEJ’s IT department built “Seven Central”—a new platform for practical data use launched in 2020 to support the company’s future IT strategies and digital transformation initiatives.

At its core, Seven Central’s ultimate purpose is to allow real-time data views. Versatile, real-time datasets—such as point-of-sale (POS) data from 7-Eleven stores—are consolidated into a centralized location in the cloud. They created a simple data mart that provides data via an API to enable them to respond more quickly to requests from individual departments. 

“In such uncertain times, it’s vital to use data to make quick decisions,” says Izuru Nishimura, Executive Officer and Head of ICT Department. “Each department across the entire company will be able to gain an immediate understanding of the situation based on the most up-to-date data and respond accordingly. This is why we built Seven Central.”

Google Cloud selected to help SEJ build and grow their data cloud

Today’s rapidly changing business environment has also highlighted the risk of IT support becoming a bottleneck. The long-term strategy is to gradually expand the datasets managed and collected in Seven Central according to business needs. 

In the first phase, SEJ collected POS data from all 21,000+ stores to enable real-time analysis. Moving forward, they would like to collect other relevant data—for example, unstructured data, such as images and videos, or master datasets that are currently stored externally. 

Google Cloud was already a top contender when SEJ started developing Seven Central in 2019. They compared various public cloud services besides Google Cloud, focusing on three main capabilities.

“We placed particular emphasis on service scalability to drive future digital transformation; security when handling data, which is the lifeline of our company; and finally, openness,” says Nishimura. He emphasizes that openness was perhaps the most important factor for choosing Google Cloud. Breaking away from the negative aspects of an entirely vendor-dependent system enabled them to build an agile development system with multiple vendors. 

Google Cloud technologies including BigQuery and API management platform, Apigee, play a vital role in Seven Central. BigQuery’s high-speed processing at petabyte scale and fully managed infrastructure helped keep costs low during development and verification.   

“Data is stored in a way that allows you to share it easily across organizations, which helps solve the issue of data silos from the perspective of scalability. I also like the fact there are some interesting features that could be used in the future—like BigQuery ML, which enables machine learning on BigQuery,” says Nishimura. 

Apigee allows SEJ to separate datasets and business logic, which is one of the key points of Seven Central. While the trend these days is to standardize interfaces using an API, the reality tends to involve many different APIs rather than the introduction of one unified API. With Apigee, SEJ provides a single unified API for all of its data cloud, and they can now understand what data is used thanks to Apigee’s API usage visualizations.  

“Right now, we collect data from all 21,000+ stores,” says Nishimura. “But in anticipation of a future expansion in business operations, we have designed a system that can scale up and run without issue, even if we were to have 30,000 stores, with 1,000 customers per store per day, purchasing five items per person.”

Real-time insights with BigQuery and Cloud Spanner

Google Cloud partner Cloud Ace came on board early in the planning phases. Based on their recommendations, SEJ decided to continue making full use of BigQuery to analyze data collected from all 21,000+ stores throughout Japan, while also using Cloud Spanner’s availability, near-unlimited scalability and transactional consistency to help achieve the real-time results needed for the project.

“Given that both the data and the regularity with which it is accessed are expected to steadily increase in the future, we chose Cloud Spanner as backend storage for data delivery via API. We consider it a good choice,” says Shota Kikuchi, General Manager, Consulting Department, Technology Division, Cloud Ace Co., Ltd.

Finally, they chose to use Google Cloud’s Stream Analytics Solutions messaging service for collecting POS data in real time, which can then be put to immediate use with Cloud Spanner and BigQuery. 

High-speed responses exceed targets and create new value 

Seven Central went live in September 2020 with surprising results. 

They initially set a target time of one hour from when a customer makes a purchase to the point when Seven Central can use that data. But when the final system was first tried—it took barely a minute. Moving forward they estimate that the latest inventory data from the service side will become available within a few minutes of being added to the system.  

“This is real innovation, and I must admit that I am quite surprised. As well as being able to solve existing issues, we also hope it will lead to new improvements and services that have been unimaginable up until now,” says Nishimura.

The team hopes to roll out the Seven Central platform in all companies affiliated with Seven & i Holdings—not just SEJ. They also plan to explore Google Cloud AI and machine learning technologies to take on challenges in new areas. For example, they are investigating the idea of clustering individual stores using BigQuery ML.

Seven Central has already attracted attention from many departments and received a lot of requests. Nishimura and his team say they hope to continue to grow Seven Central while still observing their fundamental principles—not including business logic, maintaining real-time results, and staying true to the uniqueness of SEJ.

Learn more about Google Cloud smart analytics solutions.

This post shows how to build an image search utility using natural language queries. Our aim is to use different GCP services to demonstrate this. At the core of our project is OpenAI’s CLIP model. It makes use of two encoders – one for images and one for texts. Each encoder is trained to learn representations such that similar images and text embeddings are projected as close as possible.

We will first create a Flask-based REST API capable of handling natural language queries and matching them against relevant images. We will then demonstrate the use of the API through a Flutter-based web and mobile application. Figure 1 shows how our final application would look like:

All the code shown in this post is available as a GitHub repository. Let’s dive in. 

Application at a high-level

Our application will take two queries from the user:

• Tag or keyword query. This is needed in order to pull a set of images of interest from Pixabay. You can use any other image repositories for this purpose. But we found Pixabay’s API to be easier to work with. We will cache these images to optimize the user experience. Suppose we wanted to find images that are similar to this query: “horses amidst flowers”. For this, we’d first pull in a few “horse” images and then run another utility to find out the images that best match our query.  
• Longer or semantic query that we will use to retrieve the images from the pool created in the step above. These images should be semantically similar to this query. 

Note: Instead of two queries, we could have only taken a single long query and run named-entity extraction to determine the most likely important keywords to run the initial search with. For this post, we won’t be using this approach. 

Figure 2 below depicts the architecture design of our application and the technical stack used for each of the components.

Figure 2 also presents the core logic of the API we will develop in bits and pieces in this post. We will deploy this API on a Kubernetes cluster using the Google Kubernetes Engine (GKE). The following presents a brief directory structure of our application code-base:

Next, we will walk through the code and other related components for building our image search API. For various machine learning-related utilities, we will be using PyTorch. 

Building the backend API with Flask

First, we’d need to fetch a set of images with respect to user-provided tags/keywords before performing the natural language image search. The utility below from the pixabay_utils.py script can do this for us:

  def fetch_images_tag(pixabay_search_keyword, num_images):
    """
    Fetches images from Pixabay w.r.t a keyword.
    :param pixabay_search_keyword: Keyword to perform the search on Pixabay.
    :param num_images: Number of images to retrieve.
    :return: List of PIL images.
    :return: List of image URLs.
    """
    query = (
        PIXABAY_API
        + "&q="
        + pixabay_search_keyword.lower()
        + "&image_type=photo&safesearch=true&per_page="
        + str(num_images)
    )
    
    response = requests.get(query)
    output = response.json()

    all_images = []
    all_image_urls = []

    for each in output["hits"]:
        imageurl = each["webformatURL"]
        response = requests.get(imageurl)
        image = Image.open(BytesIO(response.content)).convert("RGB")
        all_images.append(image)
        all_image_urls.append(imageurl)

    return (all_images, all_image_urls)

Note that all the API utilities are logging relevant information. But for brevity, we have omitted the lines of code responsible for that. Next, we will see how to invoke the CLIP model and select the images that would best match a given query semantically.  For this, we’ll be using Hugging Face, an easy-to-use Python library offering state-of-the-art NLP capabilities. We’ll collate all the logic related to this search inside a SimilarityUtil class:

  class SimilarityUtil:
    def __init__(self):
        self.model = CLIPModel.from_pretrained(CLIP_MODEL)
        self.processor = CLIPProcessor.from_pretrained(CLIP_PREPROCESSOR)
        self.device = "cuda" if torch.cuda.is_available() else "cpu"

    def perform_sim_search(self, images, query_phrase, top_k=3):
        """
        Performs similarity search between the images and query.
        :param images: A list of PIL images initially retrieved with
        respect to some entity e.g. Tiger.
        :param query_phrase: A list containing a single text query,
        e.g. "Tiger drinking water".
        :param top_k: Number of top images to return from `images`.
        :return: Top-k indices matching the query semantically and
        their similarity scores.
        """
        model = self.model.to(self.device)
        # Obtain the text-image similarity scores
        with torch.no_grad():
            inputs = self.processor(
                text=[query_phrase], images=images, return_tensors="pt", padding=True
            )
            inputs = inputs.to(self.device)
            outputs = model(**inputs)

        # Image-text similarity scores
        logits_per_image = outputs.logits_per_image.cpu()
        (top_indices, top_scores) = self.sort_scores(logits_per_image, top_k)

        return (top_indices, top_scores)

    def sort_scores(self, scores, top_k):
        """
        Sorts the scores in a descending manner.
        :param scores: Scores to sort through.
        :param top_k: Number of top scores to return.
        :return: Top-k scores and their indices.
        """
        values, indices = scores.squeeze().topk(top_k)
        top_indices, top_scores = [], []

        for score, index in zip(values, indices):
            top_indices.append(int(index.numpy()))
            score = score.numpy().tolist()
            top_scores.append(round(score, 3))

        return (top_indices, top_scores)

CLIP_MODEL uses a ViT-base model to encode the images for generating meaningful embeddings with respect to the provided query. The text-based query is also encoded using A Transformers-based model for generating the embeddings. These two embeddings are matched with one another during inference. To know more about the particular methods we are using for the CLIP model please refer to this documentation from Hugging Face. 

In the code above, we are first invoking the CLIP model with images and the natural language query. This gives us a vector (logits_per_image) that contains the similarity scores between each of the images and the query. We then sort the vector in a descending manner. Note that we are initializing the CLIP model while instantiating the SimilarityUtil to save us the model loading time. This is the meat of our application and we have tackled it already. If you want to interact with this utility in a live manner you can check out this Colab Notebook. 

Now, we need to collate our utilities for fetching images from Pixabay and for performing the natural language image search inside a single script – perform_search.py. Following is the main class of that script:

  class Searcher:
    def __init__(self):
        self.similarity_model = SimilarityUtil()

    def get_similar_images(self, keyword, semantic_query, pixabay_max, top_k):
        """
        Finds semantically similar images.
        :param keyword: Keyword to search with on Pixabay.
        :param semantic_query: Query to find semantically similar images retrieved from Pixabay.
        :param pixabay_max: Number of maximum images to retrieve from Pixabay.
        :param top_k: Top-k images to return.
        :return: Tuple of top_k URLs and the similarity scores of the images present inside the URLs.
        """
        images_redis_key = keyword + "_images"
        urls_redis_key = keyword + "_urls"

        if redis_client.exists(images_redis_key) and redis_client.exists(
            urls_redis_key
        ):
            keyword_images = redis_client.get(images_redis_key)
            keyword_image_urls = redis_client.get(urls_redis_key)
        else:
            (keyword_images, keyword_image_urls) = fetch_images_tag(
                keyword, pixabay_max
            )
            redis_client.set(images_redis_key, keyword_images)
            redis_client.set(urls_redis_key, keyword_image_urls)

        (top_indices, top_scores) = self.similarity_model.perform_sim_search(
            keyword_images, semantic_query, top_k
        )

        top_urls = [keyword_image_urls[index] for index in top_indices]

        return (top_urls, top_scores)

Here, we are just calling the utilities we had previously developed to return the URLs of the most similar images and their scores. What is even more important here is the caching capability. For that, we combined GCP’s MemoryStore and a Python library called direct-redis. More on setting up MemoryStore later. 

MemoryStore provides a fully managed and low-cost platform for hosting Redis instances. Redis databases are in memory and light-weight making them an ideal candidate for caching. In the code above, we are caching the images fetched from Pixabay and their URLs. So, in the event of a cache hit, we won’t need to call the CLIP model and this will tremendously improve the response time of our API. 

Other options for caching

We can cache other elements of our application. For example, the natural language query. When searching through the cached entries to determine if it’s a cache hit, we can compare two queries for semantic similarity and return results accordingly. 

Consider that a user had entered the following natural language query: “mountains with dark skies”. After performing the search, we’d cache the embeddings of this query. Now, consider that another user entered another query: “mountains with gloomy ambiance”. We’d compute its embeddings and run a similarity search with the cached embeddings. We’d then compare the similarity scores with respect to a threshold and parse the most similar queries and their corresponding results. In case of a cache miss, we’d just call the image search utilities we developed above. 

When working on real-time applications we often need to consider these different aspects and decide what enhances the user experience and maximizes business at the same time. 

All that’s left now for the backend is our Flask application – main.py:

  @app.route("/search", methods=["GET"])
def get_images():
    tag = request.args.get("t").lower()
    query = request.args.get("s_query").lower()
    top_k = request.args.get("k")

    (top_urls, top_scores) = searcher.get_similar_images(
        tag, query, MAX_PIXABAY_SEARCH, int(top_k)
    )

    return jsonify({"top_urls": top_urls, "top_scores": top_scores})

Here we are first parsing the query parameters from the request payload of our search API.  We are then just calling the appropriate function from perform_search.py to handle the request. This Flask application is also capable of handling CORS. We do this via the flask_cors library:

  cors = CORS(
    app,
    resources={
        r"/search/*": {"origin": "*"},
        r"/test/*": {"origin": "*"},
    },
)

And this is it! Our API is now ready for deployment. 

Deployment with Compute Engine and GKE

The reason why we wanted to deploy our API on Kubernetes is because of the flexibility Kubernetes offers for managing deployments. When operating at scale, auto scalability and load balancing are very important. With the comes the requirement of security — we’d not want to expose the utilities for interacting with any internal services such as databases. With Kubernetes, we can achieve all these easily and efficiently. 

GKE provides secured and fully managed functionalities for operationalizing Kubernetes clusters. Here are the steps to deploy the API on GKE at a glance:

• We first build a Docker image for our API and then push it to the Google Container Registry (GCR).
• We then create a Kubernetes cluster on GKE and initialize a deployment.
• We then add scalability options.
• If any public exposure is needed for the API, we then tackle it. 

We can assimilate all the above into a shell script – k8s_deploy.sh:

  ## Docker build and push ## 
# We are inside the `server` directory
docker build -t gcr.io/${PROJECT_ID}/search_service .
docker push gcr.io/${PROJECT_ID}/search_service

## Deploy on a GKE cluster ##
# Configure the Docker command-line tool to authenticate to Container Registry
gcloud auth configure-docker

# Create a Kubernetes cluster
gcloud container clusters create image-search-nlp 

# Create a deployment
kubectl create deployment clip-search --image=gcr.io/${PROJECT_ID}/search_service

# Number of worker replicas
kubectl scale deployment clip-search --replicas=3

# HorizontalPodAutoscaler resource
kubectl autoscale deployment clip-search --cpu-percent=80 --min=1 --max=5

# Expose deployment
kubectl  expose deployment clip-search --name=clip-search-service --type=LoadBalancer --port 80 --target-port 8080

These steps are well explained in this tutorial that you might want to refer to for more details. We can configure all the dependencies on our local machine and execute the shell script above. We can also use the GCP Console to execute it since a terminal on the GCP Console is pre-configured with the system-level dependencies we’d need. In reality, the Kubernetes cluster should only be created once and different deployment versions should be created under it. 

After the above shell script is run successfully, we can run kubectl get service to know the external IP address of the service we just deployed:

  NAME                       CLUSTER-IP    EXTERNAL-IP     PORT(S)          AGE
clip-search-service      10.3.251.122    203.0.113.0     80:30877/TCP     10s

We can now consume this API with the following base URI: http://203.0.113.0/. If we wanted to deal with only http-based API requests, then we are done here. But secured communication is often a requirement in order for applications to operate reliably. In the following section, we are to discuss how to configure the additional items to allow our Kubernetes cluster to allow https requests.  

Configurations for handling https requests with GKE

A secure connection is almost often a  must-have requirement in modern client/server applications. The front-end Flutter application would be hosted on GitHub Pages for this project, and it requires https-based connection as well. Even if configuring https connection particularly for a GKE-based cluster can be considered a chore, its setup might seem daunting at first.

There are six steps to configure https connection in the GKE environment: 

1. You need to have a domain name, and there are a lot of inexpensive options that you can buy. For instance, mlgde.com domain for this project is acquired via Gabia which is a Korean service provider.
2. A reserved (static) external IP address has to be acquired via gcloud command or GCP console. 
3. You need to bind the domain name with the acquired external IP address. This is a platform-specific configuration that issued the domain name to you. 
4. There is a special ManagedCertificate resource which is specific to the GKE environment. ManagedCertificate resource specifies the domain that the SSL certificate will be created for, so you need this. 
5. An Ingress resource should be created by listing the static external IP address, ManagedCertificate resource, and the service name and port which the incoming traffic will be routed to. The Service resource could remain the same as in the above section with only changes from LoadBalancer to ClusterIP. 
6. Last but not least, you need to modify the existing Flask application and Deployment resource to support liveness and readiness probes which are used to check the health status of the Deployment. The Flask application side can be simply modified with the flask-healthz Python package, and you only need to add livenessProbe and readinessProbe sections in the Deployment resource. In the code example below, the livenessProbe and readinessProbe are checked via /alive and /ready endpoints respectively.

  from flask_healthz import healthz
from flask_healthz import HealthError


app = Flask(__name__)
app.register_blueprint(healthz, url_prefix="/")


def printok():
   print("Everything is fine")


def liveness():
   try:
       printok()
   except Exception:
       raise HealthError("Can't connect to the file")


def readiness():
   try:
       printok()
   except Exception:
       raise HealthError("Can't connect to the file")


app.config.update(
   HEALTHZ = {
       "alive": "main.liveness",
       "ready": "main.readiness",
   }
)

One thing to be careful of is the initialDelaySeconds attribute of the probes. It is uncommon to configure this attribute with a big number, but it could be bigger than 90 – 120 seconds depending on the size of the model to be used. For this project, it is configured in 90 seconds in order to wait until the CLIP model is fully loaded into memory (full YAML script here).

  livenessProbe:
 httpGet:
   path: /alive
   port: 8080
 initialDelaySeconds: 90
 periodSeconds: 10

readinessProbe:
 httpGet:
   path: /ready
   port: 8080
 initialDelaySeconds: 90
 periodSeconds: 10

Again, these steps may seem daunting at first, but it will become clear when you have done it once. Here is the official document for Using Google-managed SSL certificates You can find all the GKE-related resources used in this project here. 

Once every step is completed you should be able to see your server application running on the GKE environment. Please make sure to run kubectl apply command whenever you create Kubernetes resources such as Deployment, Service, Ingress, and ManagedCertificate, and it is important to wait for more than 10 minutes until the ManagedCertifcate provisioning is done. 

You can run gcloud compute addresses list command to find out the static external IP address that you have configured.

  NAME  ADDRESS/RANGE  TYPE      PURPOSE  NETWORK  REGION  SUBNET  STATUS
gde      34.149.231.34       EXTERNAL                                                                    IN_USE

Then, the IP address has to be mapped to the domain. Figure 3 is a screenshot of a dashboard from where we got the mlgde.com domain. It clearly shows mlgde.com is mapped to the static external IP address configured in GCP.

In case you’re wondering why we didn’t deploy this application on App Engine, well that is because of the compute needed to execute the CLIP model. App Engine instance won’t fit in that regime. We could have also incorporated compute-heavy capabilities via a VPC Connector. That is a design choice that you and your team would need to consider. In our experiments, we found the GKE deployment to be easier and suitable for our needs. 

Infrastructure for the CLIP model

As mentioned earlier, at the core of our application is the CLIP model. It is computationally a bit more expensive than the regular deep learning models. This is why it makes sense to have the hardware infrastructure set up accordingly to execute it. We ran a small benchmark in order to see how a GPU-based environment could be beneficial here. 

We ran the CLIP on a Tesla P100-based machine and also on a standard CPU-only machine 1000 times. The code snippet below is the meat of what we executed:

  DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

start_time = time.time()
for _ in range(1000):
    with torch.no_grad():
        model = model.to(DEVICE)
        inputs = processor(text=[semantic_search_phrase], 
                        images=all_images, return_tensors="pt", padding=True)
        inputs = inputs.to(DEVICE)
        outputs = model(**inputs)
end_time = time.time() - start_time
print(f"Total time: {end_time:.3f} seconds.")

As somewhat expected, with the GPU, the code took 13 minutes to complete execution. With no GPU, it took about 157 minutes.

It is uncommon to leverage GPUs for model prediction because of cost restrictions, but sometimes we have to access GPUs for deploying a big model like CLIP. We configured a GPU-based cluster on GKE and compared the performance differences with and without it. It took about 1 second to handle a request with GPU and MemoryStore cache while it took more than 4 seconds with MemoryStore only (without the GPUs). 

For the purposes of this post, we used a CPU-based cluster on Kubernetes. But It is easy to configure GPU usage in a GKE cluster. This document shows you how to do so. For a short summary, there are two steps. First, a node should be configured with GPUs when creating a GKE cluster. Second, GPU drivers should be installed in GKE nodes. You don’t need to visit and manually install GPU drivers for each node by yourself. Rather you can simply apply the DaemonSet resource to GKE as described here.

Setting up MemoryStore

In this project, we first query the general concept of images to Pixabay, then we filter the images with a semantic query using CLIP. It means we can cache the initially retrieved images from Pixabay for the next specific semantic query. For instance, you may want to search with “gentleman wearing tie” at first, then you may want to retry searching for “gentleman wearing glass”. In this case, the base images remain all the same, so they could be stored in a cache server like Redis. 

MemoryStore is a GCP service wrapping the Redis which is an in-memory data store, so you can simply use a standard Redis Python package for accessing it. The only thing to be careful about when provisioning a MemoryStore Redis instance is to make sure it is in the same region where your GKE cluster or Compute Engine instance is.

The code snippet below shows how to make a connection to the Redis instance in Python. Nothing specific to GCP, but you only need to be aware of the usage of the standard redis-py package. 

  # REDISHOST is the IP address to the MemoryStore instance
redis_host = os.environ.get("REDISHOST", "localhost")
redis_port = int(os.environ.get("REDISPORT", 6379))
redis_client = redis.StrictRedis(host=redis_host, port=redis_port)

After creating a connection, you can store and retrieve data from MemoryStore. There are more advanced use cases of Redis, but we only used exists, get, and set methods for the demonstration purpose. These methods should be very familiar if you know maps, dictionaries, or other similar data structures. For the code portion that uses Redis-related utilities, please refer to the Searcher Python class we discussed in an earlier section. 

In the URLs below, you can find side-by-side comparisons of using MemoryStore:

• Without MemoryStore: https://youtu.be/7B88Eyrd-4s
• With MemoryStore (1st try): https://youtu.be/LE6xeEIRuMM
• With MemoryStore (2nd try): https://youtu.be/rRfK17sdk84 

Putting everything together

All that’s left now is to collate the different components we developed in the sections above and deploy our application with a frontend. All the frontend-related code is present here. 

• The front-end application is written in the Flutter development kit. The main screen contains two text fields for queries to Pixabay and CLIP model respectively. When you click the “Send Query” button, it will send out a RestAPI request to the server. After receiving the result back from the server, the retrieved images from the semantic query will be displayed at the bottom section of the screen. 
  • Please note that a Flutter application can be deployed to various environments including desktop, web, iOS, and Android. In order to keep as simple as possible, we chose to deploy the application to the GitHub Pages. Whenever there is any change to a client-side source directory, the GitHub Action will be triggered to build a web page and deploy the latest version to the GitHub Pages. 

Our final application is deployed here and it looks like so:

Note that due to constraints, the above-mentioned URL will only be live for one or two months.  

It is also possible to redeploy the back-end application with a GitHub Action. 

• The very first step is to craft a Dockerfile like below. Since Python is a scripting language, and there are lots of heavy packages that the application is dependent on, it is important to cache the steps. For instance, installing the dependencies should be separated from other commands.

  FROM pytorch/pytorch:latest
WORKDIR /app

# install the dependencies
COPY requirements.txt requirements.txt
RUN pip install -r requirements.txt

COPY . .  

# set up environment variables
ENV PIXABAY_API_KEY="..."
ENV REDIS_IN_USE="true"
ENV REDISHOST="..."

# expose port that Flask app is listening on
EXPOSE 8080

# run the Flask app
CMD [ "python3", "main.py" ]

• With the Dockerfile defined, we can use a GitHub Action like this for automatic deployment. 

Edge cases

Since the CLIP model is pre-trained on a large corpus of image and text pairs it’s likely that it may not generalize well to every natural language query we throw at it. Also, because we are limiting the number of images on which the CLIP model can operate, this somehow restricts the expressivity of the model.

We may be able to improve the performance for the second situation by increasing the number of images to be pre-fetched and by indexing them into a low-cost and high-performance database like Datastore. 

Costs

In this section, we wanted to provide the readers a breakdown of the costs they might incur in order to consume the various services used throughout the application. 

• Frontend hosting
  • The front-end application is hosted on GitHub Pages, so there is no expenditure for this.
• Compute Engine
  • With an e2-standard-2 instance type without GPUs, the cost is around $48.92 per month. In case you want to add a GPU (NVIDIA K80), the cost goes up to $229.95 per month.
• MemoryStore
  • The cost for MemoryStore depends on the size. With 1GB of space, the cost is around $35.77 per month, and whenever you add more GBs the cost will be doubled.
• Google Kubernetes Engine
  • The monthly cost for a 3 node GKE cluster with n2-standard-2 (vCPUs: 2, RAM: 8GB without GPUs) is about $170.19. If you add one GPU (NVIDIA K80) to the cluster, the cost goes up to $835.48.

While you may think that is a lot cost-wise, it is good to know that Google gives away free $300 credits when you create a new GCP account. It is still not enough for leveraging GPUs, but it is enough to learn and experiment with GKE and MemoryStore usage.

Conclusion

In this post, we walked through the components needed to build a basic image search utility for natural language queries. We discussed how these different components are connected to each other. Our image search API is able to utilize caching and was deployed on a Kubernetes cluster using GKE. These elements are essential when building a similar service to cater to a much bigger workload. We hope this post will serve as a good starting point for that purpose. Below are some references on similar areas of work that you can explore:

• Building a real-time embeddings similarity matching system
• Detecting image similarity using Spark, LSH and TensorFlow

Acknowledgments: We are grateful to the Google Developers Experts program for supporting us with GCP credits. Thanks to Karl Weinmeister and Soonson Kwon of Google for reviewing the initial draft of this post. 

API security incidents are increasingly common and disruptive. With the growth of API traffic, enterprises across the world are also experiencing an uptick in malicious API attacks, making API security a heightened priority. According to our latest API Security Research Report, 50% of organizations surveyed have  experienced an API security incident in the past 12 months and of those, 77% delayed the rollout of a new service or application. 

At the RSA Conference 2023 today, we’re making it faster and easier to help detect API abuse incidents with the introduction of Advanced API Security Machine Learning powered abuse-detection dashboards. Our newly introduced Machine Learning models are trained to detect business logic attacks. 

These types of attacks are notoriously hard to identify, and target APIs tied to intellectual property, business processes, or sensitive information, such as user data, listing of goods, or crediting accounts. These APIs must be accessible to provide business value, but have also become targets for attackers.

API security incidents can have a considerable impact on an organization’s operations and its bottom line. In June 2022, Imperva released a report titled Quantifying the Cost of API Insecurity, which estimates that lack of secure APIs could result in an average annual API-related total global cyber loss of between $41 billion to $75 billion annually. Furthermore, according to IBM’s  2022 Cost of a Data Breach Report, the average cost of a data breach is $4.35 million. It’s vital that organizations detect and mitigate API abuse incidents early to prevent prolonged fiscal and reputational damage to the business.

However, business logic attacks are harder to detect using static security policies, which allows attackers to manipulate legitimate functionality to achieve a malicious goal without triggering any static security alerts. For example, if a malicious actor gains control of a server and makes subtle changes, the shift in activity patterns of the server is generally undetectable to most monitoring tools. However, in this scenario, the Advanced API Security’s ML-powered API abuse detection model can help differentiate between legitimate and deviant traffic and immediately notify key stakeholders to act quickly and minimize blast radius of the problem.

The ML models that power API abuse detection have been trained and used by Google’s internal teams to help protect our public-facing APIs. The models rely on years of learning and best practices and are now available to all Apigee Advanced API Security customers.

Another challenge in detecting API abuse incidents is the volume of alerts. To reduce the risk of missing key security incidents, many static rules that detect less sophisticated attacks are incredibly sensitive: They generate a high volume of alerts. This makes finding the critical incidents within API traffic and acting to resolve them like “finding a needle in a haystack” for many IT teams. Apigee Advanced Security’s ML-powered dashboards more accurately identify critical API abuses and find similar patterns within the large number of bot alerts to help reduce the time to find and act on most important incidents.

With the help of Apigee Advanced API Security’s ML-powered abuse detection dashboards, customers can uncover critical API abuse incidents, including business logic attacks, scraping, and anomalies, faster. Critical threats are surfaced with clear and concise descriptions to capture the essence of the attack along with the most important characteristics such as the source of the attack, the number of API calls, and the duration of the attack, to help resolve the incident more rapidly. 

Machine Learning powered abuse-detection dashboards are available in Advanced API Security, a feature of Apigee API management that enables you to more easily detect API security misconfigurations, bad bots, and malicious activities. 

To get started with Advanced API Security’s ML-powered dashboards, start your free Apigee trial now.

At Google Cloud, we’re committed to making artificial intelligence (AI) accessible to everyone and easier to harness for new use cases.  That’s why we’re excited to announce the general availability of our intuitive, new visual user interface for Google Cloud’s Speech-to-Text (STT) API, right in Google Cloud Console, which makes the API much simpler and easier for developers to use.  

The STT API lets developers convert speech into text by leveraging Google’s years of research in automatic speech recognition and transcription technology. As advancements in AI continue to bring speech to new interfaces and devices, the STT API helps developers add speech functionality to their applications in order to better meet consumer demands. 

The STT API covers a wide variety of use cases, from dictation and short commands, to captioning and subtitles. Getting the most of STT, however, can be a complicated process. To achieve the highest accuracy on any AI use case requires careful testing and tuning. 

Previously, developers building on the STT API had to do this work manually by carefully experimenting with our API. Just to get started, developers needed familiarity with GCP integration concepts and had to either build their own tools or manage various scripts and API calls to fully understand the API documentation.  These actions required cumbersome and time-consuming effort and made measuring, customizing, and improving models even more difficult.  

Today’s announcement significantly simplifies the process, facilitating iteration and integration of models into developers’ applications by letting developers perform every API function from within the Google Cloud Console.  These tools will make it easier for developers to integrate the STT API with their products or services. This update also gives developers the ability to manage and quickly iterate on their STT model customizations with Model Adaptation.

Model Adaptation allows developers to customize STT specifically for their domains or use cases. Developers can maintain lists of words and weights that will be applied to either every request or just single requests, depending on their needs. Model adaptations are reusable and composable, so once developers have seen good results in the STT Cloud Console, they can deploy to their entire solution. 

The Speech-to-Text Cloud Console and Model Adaptation API is available now in all Google Cloud regions and languages and is accessible to all GCP users with no additional cost to that of the underlying API usage. The STT API supports over 70 languages in 120 different local variants. If you’re a developer looking for an easy to use, easy to integrate, and high-quality STT experience, sign up for our free trial and try our new interface on your own datasets today!