TrustRadius: an HG Insights company

Amazon Bedrock

Score9.4 out of 10

10 Reviews and Ratings

What is Amazon Bedrock?

Amazon Bedrock offers a way to build and scale generative AI applications with foundation models, providing a developer experience to work with a broad range of FMs from AI companies like AI21 Labs, Anthropic, Cohere, Meta, Stability AI, and Amazon.

Read more details.

Categories & Use Cases

Videos

Amazon Bedrock Key Features

Learn about the best (and worst!) features Amazon Bedrock has to offer, as determined by TrustRadius' reviewers.
Based on 10 ratings of Amazon Bedrock's features
View all Amazon Bedrock features

Top Performing Features

  • Model deployment

    After training is complete, models can be integrated into business applications using API endpoints or other convenient, standard methods.

    Category average: 8.2

  • Data monitoring and version control

    Teams can track which data is used in training at which point and roll back to previous versions as needed.

    Category average: 7.2

  • Data management

    Ingested data can be stored and prepped, with structures like data lakehouses available to handle large amounts of data from disparate sources.

    Category average: 7.9

Areas for Improvement

  • Security and compliance

    End to end encryption, GDPR compliance, SSO, role-based permissioning, and other precautions are available to protect proprietary business data.

    Category average: 8.2

  • Managed scaling

    The platform provides the computing resources needed when they’re needed, allowing users to scale training and use up or down.

    Category average: 7.7

  • Automated model training

    After teams begin the process, model training can continue autonomously, enabling faster deployment.

    Category average: 7.7

Who Buys & Uses Amazon Bedrock

Amazon Bedrock Reviews

What users are saying about Amazon Bedrock.
View all Amazon Bedrock reviews

Bedrock is good at its model catalogue.

Incentivized
Verified User
Professional in Information Technology (10,001+ employees)

Use Cases and Deployment Scope

Currently, we are utilizing the Bedrock service and its model. Right now, for security purposes, we are using the AWS SecHub solution. When a new alert from AWS for Bedrock arrives, it's very hard to understand and resolve; AWS should look into simplifying it.

Pros

  • Variety of model.
  • Their ease of subscription.

Cons

  • Integrity with AWS Sechub.
  • Bedrock documentation.

Return on Investment

  • AWS bedrock should provide proper documentation.
  • Enterprise support should be improve for bedrock service.

Usability

Alternatives Considered

GitHub Copilot and Google Gemini

Other Software Used

GitHub Copilot, Google Gemini

Resource And Cost Efficient with multi models and feature under one Name i.e. Amazon Bedrock

Incentivized
Verified User
Manager in Information Technology (51-200 employees)

Use Cases and Deployment Scope

We used Amazon Bedrock for creating agents over its AI models like Sonnet, Opus, Mistral etc. There are multiple use cases where I used it like Meeting Summary Generator, Credit Access Memo Generator based on provided information for Venture firm, Fee Management Extractor from any document. We also created Retrieval Augmented Generator for CAMs that we're manually created and One of the agent use to extract data for creating new CAM from historical information. Apart from this we're now planning to build sales forecasting model using Amazon Bedrock Agents and APIs. Amazon Bedrock provides most of things under one umbrella that we need for creating AI based systems like Models, Knowledge Bases, Guardrails etc.

Pros

  • It provides almost all famous models under one service and easy to switch between them as per requirement and testings. Models include LLM as well as embedding models
  • With few clicks yuou can deploy your own RAG using knowledge base, embedding model and LLM models.
  • If you're using any third party vector database service then you can attach it's billing/invoice with your amazon monthly invoice using Amazon market place.
  • You can set the Guard rails to avoid explicit answer and restricting model to answer on particular tops only
  • You can use models with most of the programming language using APIs like boto3 in python.

Cons

  • As of now model automatic fallback is not provided in Amazon Bedrock. Would be helpful if this is provided. As of now this needs to be handle through coding.
  • If someone asks same question repeatedly then tokens are consumed for every request. Better to have some caching for this.

Return on Investment

  • With Amazon Bedrock you don't need to keep you server running all the time and paying for it even if it is sitting ideal. This is pay per use and on demand service which saves lot in terms of money.

Usability

Alternatives Considered

Gemini Enterprise Agent Platform and Azure OpenAI Service

Amazon Bedrock Alternatives

Products similar to Amazon Bedrock that may also meet your needs.
View all Amazon Bedrock alternatives
OpenAI API Platform
CompareLearn more
Azure OpenAI Service
CompareLearn more
Cohere Command
CompareLearn more
Gemini Enterprise Agent Platform
CompareLearn more
IBM watsonx.ai
CompareLearn more