Skilled workers are in high demand across healthcare, manufacturing, agriculture, financial services, and other industries. But as much of the “baby boomer” generation — comprising 70 million people — faces retirement, companies are struggling to hire enough employees to meet business demands. A tighter labor market can also increase the risks of worker burnout and higher turnover.
As we head into a new year, we’re entering a new phase of agentic AI to help industries scale and solve more complex challenges.
Companies across many industries are turning to customized AI agents built on their own intellectual property to help employees boost productivity, develop new skills, and work smarter. These agents will become a core focus for AI development in 2025. Some are already available.
For instance, SAP offers a generative AI agent, called Joule, that is designed to deliver tailored content specific to different roles, streamline processes for deeper insights, and save time while boosting efficiency.
In the trillion-dollar healthcare industry, hospitals and clinics are turning to AI agents to help the global nursing shortage. These software agents are designed to support business- or industry-specific tasks, like taking vital signs, scheduling appointments, keeping patient records up to date, or, in the case of Hippocratic AI, proactively engaging patients over the phone on a wide range of non-diagnostic tasks, from checking in to answer their questions pre-procedure to checking in and reminding them to follow their treatment plans afterward.
Putting AI to work: Custom AI models are transforming industries
Hundreds of systems integrators are delivering a broad range of customized AI agents to ease tasks for workers so they can focus on solving businesses' toughest challenges. Industries including banking, retail, and software and information services are projected to allocate nearly $90 billion toward AI this year, representing 38% of the global AI market, according to IDC1.
The trillion-dollar telecommunications industry helps businesses drive efficiency and increase customer satisfaction and employee productivity. Amdocs, a provider of software and services for communications and media providers, has developed a generative AI agent as part of its amAIz platform. Amdocs’ amAiz Agents help reduce tokens consumed in data preprocessing by 60% and tokens consumed in inferencing by 40%. It also lowers query latency by nearly 80% and improves accuracy of responses by up to 30%.
Similarly, BT Group piloted ServiceNow’s generative AI experience Now Assist for Telecom Service Management, reducing the amount of time customer support agents spent on tasks such as writing case summaries and reviewing complex notes by 55%.
In financial services, European neobank bunq has developed a personal AI assistant called Finn to benefit customers and employees alike. Finn handles more than half of bunq’s automated tickets and can spot fake methods of identification when onboarding the digital bank’s new users.
Cybersecurity, another billion-dollar industry, has seen a 600% rise in cybercrime over the last four years. To prevent and combat these threats, the industry is using AI agents to enhance security measures and streamline operations. Global cybersecurity company Trend Micro has developed an AI-powered assistant that provides proactive threat detection and automated incident response, enabling organizations to swiftly mitigate risks and improve overall security capabilities.
Generative AI startup Abridge, is focused on unburdening clinicians from clerical work so they can focus on what matters most - their patients. Abridge’s enterprise-grade AI technology transforms patient-clinician conversations into structured clinical notes in real-time with deep EMR integrations. With support for multiple languages and specialties, Abridge supports a wide range of clinician and patient encounters.
Follow the money — and efficiency
Driving the rise in custom AI agents is retrieval-augmented generation (RAG). This new technology is helping boost spending on AI to $632 billion by 2028, up from $235 billion today, according to IDC.
RAG allows AI agents to retrieve contextually relevant information from a vast database, ensuring that responses are accurate and contextually appropriate. It allows for dynamic updates, creating a data flywheel that helps provide the most up–to-date information for AI agents and enhances accuracy.
By continuously learning from user interactions and preferences, RAG-based agents can scale and adapt their responses to better align with individual user styles and requirements. This ongoing learning process enables the AI to offer more personalized communication, improving user engagement and satisfaction.
Business Value
Integrating AI into business operations is no longer a luxury but a necessity. Unlike a generic, one-size-fits-all approach, personalized agents can meet specific challenges and create business opportunities that help companies boost efficiency, improve customer satisfaction, and inform decision-making.
As artificial intelligence continues to evolve in the new year, businesses that embrace customized AI agents will be poised to lead the charge in their respective industries.
Amanda Saunders is director of AI Software at Nvidia.