IBM Introduces ‘Lightweight Engine’ for WatsonX.ai Service

IBM recently launched the ‘Lightweight Engine’ as part of its WatsonX.ai service. This new feature will initially be aimed at larger enterprises. However, it could also play an important role for smaller companies looking to scale their AI capabilities, as well as mid-market companies in fast-growing sectors such as fintech.

Generative AI market driving revenue growth

Without a doubt, the generative AI market will drive a large portion of the tech sector’s revenue growth in the first half of 2024. Just 10 years ago, hardly anyone could have predicted the tremendous growth of this sector, which is largely driven by the popularity of large language models such as OpenAI’s ChatGPT and Anthropic’s Claude.

Before the launch of ChatGPT, it was widely believed by experts in AI and finance that large language models like GPT-3 were not reliable enough for applications in finance or other error-prone domains. Despite the progress made since then, the caveat remains that AI models trained on public data can be unpredictable.

An example of generative AI being applied to finance is JPMorgan Chase, which recently acquired access to OpenAI’s ChatGPT for its 60,000 employees, integrating internal data and custom security measures. This illustrates that even the conservative financial sector is increasingly embracing generative AI.

Other AI services now also offer business options

While many popular AI services, such as ChatGPT, now offer enterprise options, these are typically entirely cloud-based. In industries such as fintech, where regulations and security require data to be isolated from external influences, cloud-based AI solutions may not always meet the requirements.

IBM’s WatsonX.ai offers both cloud-based and on-premise solutions, and with the addition of the ‘Lightweight Engine’, models can now be run locally, with a smaller footprint.

Savio Rodrigues, IBM’s vice president of ecosystem engineering and developer advocacy, explains: “Enterprises looking to add on-premises solutions are looking for the lightest platform to run their generative AI scenarios, without wasting CPUs or GPUs. WatsonX.ai’s lightweight engine enables ISVs and developers to scale GenAI solutions for enterprises while optimizing costs.”

Flexibility of this system can make the difference

In fast-growing industries like fintech, blockchain, and crypto lending, where third-party AI solutions don’t always meet stringent security requirements, the flexibility of a hybrid cloud and on-premise solution can make the difference between developing models in-house or sourcing an external service.

Although there are several competing services, IBM’s ‘Lightweight Engine’ seems to live up to its name. However, the smaller footprint and greater efficiency come at the cost of reducing the number of features available in the full version.