Pioneering Intelligence

The potential of artificial intelligence (AI) to revolutionize the public sector has been realized—delivering substantial increases in productivity and operational efficiencies and delivering innovations such as predictive healthcare, adaptive education, and optimized crisis and weather response. It has changed our ability to manage and treat health conditions quickly, more effectively, and at scale. Additionally, AI has enabled virtual assistants and apps that help citizens and businesses better navigate government services and agencies and deliver personalized services through optimized digital experiences that more closely mirror citizens’ experiences in the private sector.

As the public sector faces the next wave of technological transformation with generative AI, agencies should take steps to equip themselves to leverage this technology. Our research and experience within this space indicate three key steps agencies should consider:

Every transformation starts with a single step. The first step for artificial intelligence (AI) in the public sector was adopting AI applications to predict, simulate, and automate agency operations and mission-oriented outcomes. 

With generative AI, your next steps could be the catalyst to propel your agency into the vanguard of this technological transformation. How do you take this next step? How can you understand, adopt, and integrate generative AI into your organization? Here are three actionable steps to get started, with a special spotlight on a tool that can help power your progress — the NVIDIA NeMo framework, part of the NVIDIA AI Enterprise software platform.

STEP ONE: Set Your Course: Strategy, Mission, and Trust

Launch your AI journey by aligning AI strategy with your agency's mission and competence. This process should leverage insights from mid- and senior-level management to formulate an AI vision and identify high-priority AI use cases. These individuals can act as AI champions, fostering trust as generative AI is introduced into daily operations. Establishing a clear intention will help steer your organization in the right direction during this exciting tech revolution. 

In the rapidly evolving landscape of AI, businesses and researchers are facing the critical challenge of promoting innovation while ensuring trustworthiness. Deloitte’s Trustworthy AI Framework involves a delicate balance between advancing innovation and maintaining trustworthiness.

1. Ethical AI principles necessitate a fair and impartial approach, allowing equal treatment across all participants. 
2. Transparency and interpretability are key, ensuring that users understand the processes and decisions behind AI systems. 
3. Responsibility and accountability must be clearly delineated, providing a concrete framework for managing AI outcomes. 
4. Robustness and reliability should underscore system design, enabling consistent learning and dependable results. 
5. Privacy is paramount, respecting personal data and offering autonomy over data sharing. 
6. Lastly, AI systems should uphold safety and security, safeguarding against potential risks. 

Embracing these six principles ensures ethical AI applications, fostering a sustainable ecosystem that marries technological advancements with user trust. By embedding these principles, we can forge a path towards responsible AI use, underpinning progress with trust and integrity.

STEP TWO: Harness the Power of AI: Leverage AI Specialists and Tools to Unlock Performance 

In order to leverage the power of generative AI, organizations must bridge the knowledge gap and collaborate with AI specialists who understand the need for high-performing tools. As technology leaders and integrators capable of thriving in a rapidly changing tech ecosystem, these specialists can help provide the requisite infrastructure and development frameworks to enable generative AI capabilities.

A key component in generative AI is the underlying infrastructure. The NVIDIA NeMo Large Learning Model (LLM)  software framework is a leader in conversational AI development. With its open-source toolkit, it provides researchers and developers with a comprehensive suite of tools and components, making it easier to create top-tier conversational AI systems. NVIDIA NeMo's end-to-end model training and deployment, which prioritize speech and natural language processing tasks, supercharge progress and are available through the NVIDIA AI Enterprise software platform as a production-grade generative AI solution. NeMo is a powerful tool, and with the capability to handle trillions of operations per second (TOPS), NVIDIA GPU-accelerated computing can help provide the fuel needed to power LLMs, amplify progress, and provide results.

STEP THREE: Develop Expertise: Create, Learn, and Refine 

Begin by laying out a roadmap that directs your agency towards understanding the value and amplifying effects of generative AI:

1.  Identify suitable use cases and kick off with low-risk pilot projects before scaling up. 

2. Consider investing in pre-trained models, like those available through the NVIDIA NeMo Framework, that can accelerate learning among your core group of AI champions, bringing your organization closer to a model that better serves diverse user groups across your organization.

Identify Suitable Use Cases - Work Accelerated 

Harnessing the power of generative AI can be a valuable tool for agencies as they navigate complex, technical policies. With AI, inquiries can be synthesized into simple, plain language summaries with references to the appropriate underlying policies, reducing the time and effort it takes to research and analyze. This can foster robust research and informed decision-making, as well as improve customer experience and reduce reputational risk for the organization. For example, Deloitte teamed up with NVIDIA to develop a prototype using NVIDIA Parabricks to provide genomic sequencing of bioinformatics data. Three data scientists took nearly a year to learn how to configure, troubleshoot, and develop a prototype that provided accurate sequencing and which was nearly 66 times more performant than CPU-based processing. That same three-person team can now build out the same capability in just four weeks. Generative AI can act as a productivity multiplier, freeing up staff to focus on higher-value tasks. Ultimately, agencies can leverage the power of generative AI to improve efficiency, enhance research, and streamline customer experience.

Invest in Pre-Trained Models - Interactive Evolution

To effectively integrate LLMs within an enterprise, agencies need to consider how the LLMs will be integrated into an existing set of capabilities and/or application portfolios. This includes the integration of various data sources, data warehouses, and data lakes, as well as the maintenance and frameworks behind these systems. As the volume of data available and accessible to the LLMs increases, the accuracy and precision of deterministic actions by generative AI models will improve, leading to the evolution of more interactive conversations with users. This can be beneficial in many sectors, such as healthcare, law, and finance, where it can reduce overhead and improve response rates for claims-based processing and benefits inquiries. Additionally, it can provide citizens with the option to self-service their claims without needing human interaction in the loop.

About Deloitte

Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee, and its network of member firms, each of which is a legally separate and independent entity. Please see www.deloitte.com/about for a detailed description of the legal structure of Deloitte Touche Tohmatsu Limited and its member firms. Please see www.deloitte.com/us/about for a detailed description of the legal structure of Deloitte LLP and its subsidiaries. Certain services may not be available to attest clients under the rules and regulations of public accounting.

This publication contains general information only and Deloitte is not, by means of this publication, rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional advisor. Deloitte shall not be responsible for any loss sustained by any person who relies on this publication. 

Copyright © 2023 Deloitte Development LLC. All rights reserved.