Getting Started With MongoDB and AWS CodeWhisperer

MongoDB
November 5, 2023 | Updated: July 21, 2025

All of the features of AWS CodeWhisperer have moved to Amazon Q Developer. You can see the updated post about how to use Amazon Q here.

Amazon CodeWhisperer is trained on billions of lines of code and can generate code suggestions — ranging from snippets to full functions — in real-time, based on your comments and existing code. AI code assistants have revolutionized developers’ coding experience, but what sets Amazon CodeWhisperer apart is that MongoDB has collaborated with the AWS Data Science team, enhancing its capabilities!

At MongoDB, we are always looking to enhance the developer experience, and we've fine-tuned the AWS CodeWhisperer Foundational Models to deliver top-notch code suggestions — trained on, and tailored for, MongoDB. This gives developers of all levels the best possible experience when using AWS CodeWhisperer for MongoDB functions.

This tutorial will help you get AWS CodeWhisperer up and running in VS Code, but CodeWhisperer also works with a number of other IDEs, including IntelliJ IDEA, AWS Cloud9, AWS Lambda console, JupyterLab, and Amazon SageMaker Studio. On the Amazon CodeWhisperer site, you can find tutorials that demonstrate how to set up AWS CodeWhisperer on different IDEs, as well as other documentation.

Note: Amazon CodeWhisperer allows users to start without an AWS account because, usually, creating an AWS account requires a credit card. Currently, CodeWhisperer is free for individual users. So it’s super easy to get up and running.

Installing Amazon CodeWhisperer for VS Code

CodeWhisperer doesn’t have its own VS Code extension. It is part of a larger extension for AWS services called AWS Toolkit. AWS Toolkit is available in the VS Code extensions store.

Step 1. Open VS Code and navigate to the extensions store (bottom icon on the left panel).

Step 2. Search for AWS CodeWhisperer, and it will show up as part of the AWS Toolkit.

Screen grab showing searching for the AWS CodeWhisperer in the extensions store.

Step 3. Once found, hit Install. Next, you’ll see the full AWS Toolkit Listing.

Screen grab of the store function for hitting install.

Step 4. Once installed, you’ll need to authorize Amazon CodeWhisperer via a Builder ID to connect to your AWS developer account (or set up a new account if you don’t already have one).

Screen grab of the page for authorizing Amazon CodeWhisperer.

Using AWS CodeWhisperer

Navigating code suggestions

Screen grab of the developer tools drop down window for Amazon CodeWhisperer

With Amazon CodeWhisperer installed and running, as you enter your prompt or code, CodeWhisperer will offer inline code suggestions. If you want to keep the suggestion, use TAB to accept it. AWS CodeWhisperer may provide multiple suggestions to choose from depending on your use case. To navigate between suggestions, use the left and right arrow keys to view them, and TAB to accept.

If you don’t like the suggestions you see, keep typing (or hit ESC). The suggestions will disappear, and Amazon CodeWhisperer will generate new ones at a later point based on the additional context.

Requesting suggestions manually

You can request suggestions at any time. Use Option-C on Mac or ALT-C on Windows. After you receive suggestions, use TAB to accept and the arrow keys to navigate.

Getting the best recommendations

For best results, follow these practices.

Give Amazon CodeWhisperer something to work with. The more code your file contains, the more context CodeWhisperer has for generating recommendations.
Write descriptive comments in natural language — for example.

// Take a JSON document as a String and store it in MongoDB returning the _id

//Insert a document in a collection with a given _id and a discountLevel

Specify the libraries you prefer at the start of your file by using import statements.

// This Java class works with MongoDB sync driver.
// This class implements Connection to MongoDB and CRUD methods.

Use descriptive names for variables and functions.
Break down complex tasks into simpler tasks.

Provide feedback

As with all generative AI tools, they are forever learning and forever expanding their foundational knowledge base, and MongoDB is looking for feedback. If you are using Amazon CodeWhisperer in your MongoDB development, we’d love to hear from you.

We’ve created a special “codewhisperer” tag on our Developer Forums, and if you tag any post with this, it will be visible to our Amazon CodeWhisperer project team, and we will get right on it to help and provide feedback. If you want to see what others are doing with CodeWhisperer on our forums, the tag search link will jump you straight into all the action.

We can’t wait to see your thoughts and impressions of MongoDB and Amazon CodeWhisperer together.

← Previous

Announcing LangChain Templates for MongoDB Atlas

Since announcing the public preview of MongoDB Atlas Vector Search back in June, we’ve seen tremendous adoption by developers working to build AI-powered applications. The ability to store, index, and query vector embeddings right alongside their operational data in a single, unified platform dramatically boosts engineering velocity while keeping their technology footprint streamlined and efficient. Atlas Vector Search is used by developers as a key part of the Retrieval-Augmented Generation (RAG) pattern. RAG is used to feed LLMs with the additional data they need to ground their responses, providing outputs that are reliable, relevant, and accurate for the business. One of the key enabling technologies being used to bring external data into LLMs is LangChain. Just one example is healthcare innovator Inovaare who is building AI with MongoDB and LangChain for document classification, information extraction and enrichment, and chatbots over medical data. Now making it even easier for developers to build AI-powered apps, we are excited to announce our partnership with LangChain in the launch of LangChain Templates ! We have worked with LangChain to create a RAG walkthrough using MongoDB Atlas Vector Search and vector store. This easy-to-use template can help developers build and deploy a Chatbot application over their own proprietary data. LangChain Templates offer a reference architecture that’s easily deployable as a REST API using LangServe . We have also been working with LangChain to release the latest features of Atlas Vector Search, like the recently announced dedicated vector search aggregation stage $vectorSearch, to both the MongoDB LangChain python integration as well as the MongoDB LangChain Javascript integration . Similarly, we will continue working with LangChain to create more templates, that will allow developers to bring their ideas to production faster. If you’re building AI-powered apps on MongoDB, we’d love to hear from you. Sign up to our AI Innovators program where successful applicants receive no-cost MongoDB Atlas credits to develop apps, access to technical resources, and the opportunity to showcase your work to the broader AI community.

November 2, 2023

Next →

Scale Performance with View Support for MongoDB Atlas Search and Vector Search

We are thrilled to announce the general availability (GA) of View Support for MongoDB Atlas Search and Atlas Vector Search , available on MongoDB versions 8.0+. This new feature allows you to perform powerful pre-indexing optimizations—including Partial Indexing to filter your collections, and Document Transformation to reshape your data for peak performance. View Support for MongoDB Atlas Search helps you build more efficient, performant, and cost-effective search experiences by giving you precise control over your search strategy. Let's look at how it works. How it works in 3 simple steps At its core, View Support is powered by MongoDB views , queryable objects whose contents are defined by an aggregation pipeline on other collections or views. Getting started is straightforward: Create a view: Define a standard view using an aggregation pipeline to filter or transform your source collection. This feature is designed to support views that contain the stages $match with an $expr operator, $addFields , and $set . Note: Views with multi-collection stages like $lookup are not supported for search indexing at this time. Index the view: Build your MongoDB Atlas Search or Atlas Vector Search index directly on the view you just created. Query the view: This is the best part. You run your $search , $searchMeta , or $vectorSearch queries directly against the view itself to get results from your perfectly curated data. With this simple workflow, you can now fine-tune exactly what and how your data is indexed. The two key capabilities you can use today are Partial Indexing and Document Transformation. Figure 1. High-level architectural diagram of search index replication on a view. Search indexes perform initial sync on the collection and apply the view pipeline before saving the search index to disk storage. Index only what you need with partial indexing Often, only a subset of your data is truly relevant for search. Imagine an e-commerce catalog where only "in-stock" products should be searchable or a RAG system where only documents containing vector embeddings will be retrieved. With Partial Indexing, you can use a $match stage in your view to create a highly-focused index that: Reduces index size: Dramatically shrink the footprint of your search indexes, leading to cost savings and faster operations. Improves performance: Smaller indexes mean faster queries and index build times. Optimize your data model with document transformation Beyond filtering, you can also reshape documents for optimal search performance. Using $addFields or $set in your view, you can create a search-optimized version of your data without altering your original collection. This is perfect for: Pre-computing values: Combine a firstName and lastName into a fullName field for easier searching, or pre-calculate the number of items in an array. Supporting all data types: Convert incompatible data types, like Decimal128, into search-compatible types like Double. You can also convert booleans or ObjectIDs to strings to enable faceting. Flattening your schema: Promote important fields from deeply nested documents to the top level, simplifying queries and improving performance over expensive $elemMatch operations. For example, consider a vacation home rental company with a listings collection storing reviews as an array of objects. To enable end-users to filter for listings with > N reviews, they create a view called listingsSearchView . The view pipeline of listingsSearchView uses an $addFields stage to add the numReviews field, which is computed based on the size of the reviews array. By creating a search index on listingsSearchView , they can run efficient $search queries on numReviews without compromising data integrity in the source collection. Figure 2. High-level architectural diagram of running search queries on a view. After the search index identifies documents to return, mongod applies the view pipeline to return the view documents. Why these optimizations are critical for scaling As your application and data volume grow, search efficiency can become a bottleneck. View Support for MongoDB Atlas Search provides the critical tools you need to maintain blazing-fast performance and control costs at scale by giving you granular control over your indexes. We are incredibly excited to see how you use these new capabilities to build the next generation of powerful search experiences on MongoDB Atlas. Ready to get started? Dive into the documentation to learn more: Atlas Search , Atlas Vector Search . Note: We plan to add compatibility for more types of Views in the future. If there’s a stage that you want to see, please let us know .

August 7, 2025