AI Agents - Use Cases, Architecture and Addressable Market
Agents Are Advanced AI Systems that Will Provide Superior Scale, Reliability and Speed for Complex Consumer and Enterprise Tasks
Demand for AI and ML-powered applications is growing in the Consumer and Enterprise sectors. At the same time, the needs that they are required to serve and the tasks to accomplish are increasingly varied, which means that a single foundation model is not the most adequate solution, both from a use case and cost perspective.
That is why the concept of AI agents, which encompasses multiple models, as well as software tools, has been increasingly popular. Established software companies and start-ups are experimenting with combining multiple types of models into one system and adding tools, such as search engines, calculators and web browsers into the back-end of AI applications.
For Consumers, AI Agents will benefit a variety of use cases:
For search, RAG systems alongside LLMs will increase the reliability and factual accuracy of the results by grounding them in their source documents
For online commerce, calculators to assist with price and discount comparisons will ensure that the math is right
For mobile, where fast access to the desired functionality is important, smaller models for quick retrieval might be used instead of LLMs
For news and media, where access to recent information is essential, web crawlers and webviews embedded in AI applications will assist LLMs
For education tools, where more advanced and complex reasoning is needed, agentic systems will encompass multiple models, each responsible for a specific topic
A special segment is that of General Purpose smart assistants for consumers, which are meant to help across all of the use cases above and more. They are increasingly complex systems that will entail multiple models and software tools, in order to be able to help in all instances, across PCs, tablets and mobile devices.
We are only at the beginning of the evolution of smart assistants for consumers, which will become more advanced backed by various types and sizes of foundation models, orchestration tools and integration into popular consumer platforms, from operating systems to major websites and services.
In the Enterprise, companies are looking at AI Agents to take over increasingly complex tasks and create cost efficiencies, in order to free up human and financial resources so they can be used in more innovative ways.
For a deep dive into AI Agents’ role in the Enterprise, architectures and addressable market, as well as advice for entrepreneurs who are building them, check out the “AI Agents: Revolutionizing Enterprise Decision-Making and Complex Task Automation” article I co-authored with Tim Guleri from Sierra Ventures.