Nvidia Launches NeMo Microservices to Boost Enterprise AI Agent Development

Nvidia Introduces NeMo Microservices for Enterprise AI Development

Nvidia has announced the general availability of NeMo microservices, a comprehensive set of tools designed to accelerate the development and deployment of AI agents in enterprise environments. This release comes at a time when businesses are seeking to maximize their return on AI investments, with many reporting only minor gains so far 5

5

.

Key Components of NeMo Microservices

The NeMo microservices suite includes several key components:

1. NeMo Curator: Gathers and refines enterprise data for model training 1

   1

   2

   2

   .
2. NeMo Customizer: Applies state-of-the-art training techniques to teach models new skills and knowledge 2

   2

   3

   3

   .
3. NeMo Evaluator: Assesses model improvements and prevents regression 1

   1

   2

   2

   .
4. NeMo Guardrails: Ensures AI agents operate within intended parameters and adhere to safety and security measures 1

   1

   2

   2

   .
5. NeMo Retriever: Extracts relevant data from various sources for model use 1

   1

   3

   3

   .

These tools work together in a circular pipeline, creating what Nvidia calls a "data flywheel" to continuously improve AI models based on new data and user feedback 2

2

3

3

.

Enterprise Applications and Benefits

Nvidia's NeMo microservices are designed to integrate with existing enterprise platforms and AI software stacks. Several companies have already reported significant benefits:

• AT&T, working with Arize and Quantiphi, achieved up to 40% increase in AI agent accuracy for personalized services, fraud prevention, and network optimization 3

  3

  4

  4

  .
• Cisco's Outshift team, collaborating with Galileo, developed a coding assistant that reduced tool selection errors by 40% and improved response times by up to 10x 3

  3

  .
• Nasdaq enhanced its Gen AI Platform's search capabilities, resulting in up to 30% improved accuracy and response times 3

  3

  .

Broad Software Support and Partnerships

NeMo microservices support a wide range of popular open AI models, including Llama, Microsoft Phi, Google Gemma, Mistral, and Nvidia's own Llama Nemotron Ultra 3

3

4

4

. The toolkit has been integrated into platforms from leading AI software providers such as Cloudera, Datadog, Dataiku, and DataRobot 3

3

4

4

.

Deployment and Availability

Developers can access NeMo microservices through the Nvidia NGC catalog or as part of the Nvidia AI Enterprise suite. The tools can be deployed on various accelerated computing infrastructures, both on-premises and in the cloud, with enterprise-grade security and support 1

1

3

3

4

4

.

Industry Impact and Future Outlook

Joey Conway, Nvidia's senior director of generative AI software for enterprise, emphasized the potential of AI agents as "digital employees" that can significantly enhance productivity across various industries 1

1

. With enterprises building large-scale multi-agent systems, Nvidia positions AI agents as a trillion-dollar opportunity 3

3

.

However, research indicates that many businesses are still struggling to realize significant gains from their AI investments. A UK study found that companies are spending an average of £321,000 ($427,000) on AI for customer experience improvements, with 44% of business leaders reporting only slight improvements 5

5

.

As the AI landscape continues to evolve, Nvidia's NeMo microservices aim to address these challenges by providing enterprises with the tools to build more effective, continuously improving AI agents that can deliver tangible business value.