An AI solution might include comprehensively analyzing large amounts of detailed data to identify high-impact patterns and trends to generate a series of predicted or possible outcomes.

It can be trained to predict future resource needs based on historical data and real-time situational analysis, perhaps to recommend changes to inventory levels or production capacity.

An AI solution might include executing processes that require little human judgment or intervention.

Solutions like this allow organizations to transform business processes by automating mundane tasks and decisions, achieving higher speed and precision, increasing productivity, reducing costs, and improving customer experience.

Organizations currently scaling intelligent automation say they've already achieved significant reduction in costs on average to date.

AI "Automate" solutions typically incorporate one or more of these functions:

Capturing and preparing data

Digitizing paper documents

Processing submitted forms or applications

Using natural language processing (NLP)

Increasing data intake capacity

Displaying relevant notices, status, or balances

AI "Automate" solutions have the potential to enable end-to-end back-office automation, from service intake to feedback; increase analytics capabilities; and simplify audit and historical document tracing efforts.

How might these automation capabilities help your clients?

Deloitte has developed an AI "Detect" Solution to help the Department of Social Services (DSS) process and review 70,000+ applications per month.

The team used robotic process automation (RPA), natural language processing (NLP), and computer vision (CV) to digitize paper documents and accelerate application processing.

An AI solution might include identifying objects and patterns through the analysis of documents, pictures, and video streams.

After digitizing the data using computer vision, the solution is able to use natural language processing to enable smart searches of the information, reducing the need for manual entry and minimizing the possibility of human error.

AI "Detect" solutions typically incorporate one or more of these functions:

Digitization of paper documents

Interpretation or analysis of pictures and videos

Assessment of submitted applications

Automation of monitoring or analysis

Understanding of potential actions given certain criteria

Automation of review processes

Uncovering of suspicious activities or actors

How might these detection capabilities help your clients?

Deloitte has developed an AI solution that uses computer-vision-enabled video surveillance to detect potential security threats more quickly and accurately.

Video surveillance is an essential tool for maintaining law and order but has traditionally required constant monitoring by humans.

In prisons, AI video surveillance solutions could detect weapons and potential threats, issuing alerts and highlighting possible problems on the screen, so security personnel can quickly pinpoint and address them.

An AI "Interact" solution might include engaging with humans in real-time, two-way communication, acting as a virtual assistant. Natural language processing and machine learning are used to deliver a more efficient, engaging, and human-like customer experience, often through chatbots.

These can be trained to answer questions, schedule appointments and calls, and refer customers to the department most appropriate to handle their request.

Solutions are even being built to interpret human intent through these interactions, providing even more advanced recommendations.

Virtual Assistants can handle repetitive functions and automate information processing for agencies who tend to have lengthy call times.

It can also be used to help improve accessibility for functions by providing services in various languages for non-native English speakers.

Deloitte has developed a new-hire chatbot that allows new employees to ask questions, block calendar time, and navigate Deloitte systems and resources.

It allows employees to engage with FAQs in a more interactive, engaging format, with information provided just-in-time and within the right context.

An AI "Simulate" solution might create a virtual representation that serves as a real-time digital counterpart of a physical object or process, a technique called "digital twin."

Once a digital twin is developed, a deep learning solution can simulate tactical moves, allowing leaders to refine military strategy in real time.

Another kind of model might attempt to simulate the human decision-making of millions of individual "agents" to predict the outcomes of complex systems where each agent takes actions in their own unique ways.

AI "Simulate" models will impact several different functional areas, improving their ability to better prepare for scenarios. For government agencies charged with providing public services, models can simulate different operational scenarios to reduce wait times, improve supply chain logistic channels, and tailor services provided to cut down on administrative costs.

Simulation can be used by transportation agencies to understand how traffic patterns can change based on adding an additional lane.

In cybersecurity, simulation is used for threat modeling to better understand common patterns of behavior that can be detected to indicate malicious behavior.

FutureScape, a Deloitte modeling and simulation asset, supported a state transportation agency in identifying the riskiest roadways and intersections for drivers, bikers, and pedestrians.

This data could also be repurposed to model traffic demand patterns, which could allow users to cut down on miles traveled.

An AI "Interpret" solution might include reading documents, converting text and speech into data, and deriving language analysis from that data.

Natural language processing (NLP) and natural language generation (NLG) can summarize and normalize structured and unstructured data from varied sources to aid in analysis. Analysts can focus their efforts on interpreting findings and generating insights rather than on mundane tasks like collecting and processing text information.

The ability to have a machine interpret written text will impact multiple different types of industries.

A model that can ingest large quantities of text can generate written summaries that synthesize key points.

Interpretation can also be used to scan text and identify pre-defined key terms within large quantities of unstructured data.

Interpretation can also be used to provide additional accessibility services, where text can be translated from one language to another, while also preserving the appropriate syntax. This can help state and local agencies provide cost-effective services to an increased range of citizens.

Deloitte has developed a conversational AI solution to handle a wide range of common consumer-facing interactions—finding better credit cards, canceling unneeded accounts, or negotiating collections.

In recent years, banks have found their customer service call centers and agents overwhelmed with customer requests. Banks can relieve this pressure by using conversational AI to handle thousands of interactions with high accuracy, connecting customers to live chat only when required. The Deloitte solution helps banks provide personalized financial plans, enhance customer relationships, and even automate debt-collection activities.