SAIGroup buys the business. Then builds the AI.

Consider the hospital television. It is an unglamorous place to look for the future of artificial intelligence. Yet a business built around engaging patients, including at the bedside, became a revealing purchase for SAIGroup. In July 2024, the investment firm acquired Get Well. The attraction was easy to miss if you were looking only at algorithms: an established route into the daily life of healthcare.

Get Well had spent years helping healthcare organizations communicate with patients. SAIGroup already backed companies working on clinical intelligence. Put those capabilities together and a different possibility appears: recommendations connected to the places where patients and clinicians actually interact. That is the investment thesis in miniature. Start with a functioning business. Give it more intelligence.

A billion dollars, with a buying list

SAIGroup is the private investment and operating group founded by Romesh Wadhwani. Its public identity can be confusing: announcements also call it SymphonyAI Group, while SymphonyAI is an operating company within the portfolio. The distinction matters. SAIGroup supplies ownership, capital, and operating support; its businesses sell the software.

Wadhwani committed up to $1 billion to the group. That is a commitment, rather than a receipt for money already spent. His earlier career helps explain the appetite for established businesses: before SAIGroup, he built software investor Symphony Technology Group, and before that, Aspect Development. He had experience with both writing software and assembling companies.

The buying list is unusually concrete. For platform acquisitions, SAIGroup publishes a target of $25 million to $150 million in annual recurring revenue, with businesses around breakeven or profitable. Smaller acquisitions have a separate range. A strong client base is part of the specification. Revenue and customers arrive before the promised transformation.

“Deliver the highest value in months, not years”SAIGroup’s stated customer-focus principle

Read that as a preference for an existing commercial foothold. A software business already knows something about how its industry works: which data matters, who approves a purchase, and what users will tolerate. SAIGroup’s wager is that these assets become more useful when paired with new AI products.

The hospital deal explains the method

The healthcare sequence makes this tangible. In October 2023, SAIGroup launched RhythmX AI with $50 million in initial funding. Founded by Deepthi Bathina, it proposed a precision-care platform that would help doctors explore patient-specific recommendations through a natural-language interface. Clinical data and predictive models were central to the proposition.

Then came Get Well. The acquisition added a patient-engagement business with customers including Kaiser Permanente and Children’s Hospital Colorado. Oxford Finance supplied debt financing for the transaction. The purchase price remained undisclosed. Whatever the financial structure, the operational logic was visible: clinical recommendations and communication with patients could become parts of a more connected service.

In November 2025, Get Well and RhythmX AI announced their combination as GW RhythmX, with Bathina as CEO and Get Well founder Michael O’Neil as vice chair. The announcement described deployments across more than 150 health systems. The interesting development was organizational: two complementary businesses were being brought under one leadership structure.

That does not establish that every patient will receive better care. It does show what SAIGroup actually did. It funded a new company, bought an established one, and combined them around a shared product direction. Ownership made that sequence possible.

Different industries. Familiar paperwork.

Healthcare is only part of the portfolio. ConcertAI supplies real-world data and software for life-sciences research and clinical work. Its products address questions such as how to generate evidence and support clinical trials. In March 2022, ConcertAI raised $150 million from Sixth Street at a $1.9 billion valuation. Those numbers belong to ConcertAI, not to its parent group.

SymphonyAI addresses retail, financial services, manufacturing, media, and enterprise IT. A retailer might use its software to plan merchandise; a bank might use it to investigate financial crime. In June 2024, Microsoft recognized SymphonyAI with its Business Transformation: AI Innovation Partner of the Year award. The partnership also places Microsoft Azure inside the technology story.

The product direction keeps moving. In September 2026, SymphonyAI introduced Symphony Risk Intelligence, describing a platform for continuous compliance with governed AI agents. Its proposition is to supplement systems banks already depend on. That fits the wider pattern: work with the installed business, then change how particular tasks get done.

JazzX AI applies a related idea to residential mortgages. Its suite handles work around applications, fulfillment, and closing. Documents, missing evidence, and handoffs are the raw material. The company explicitly keeps the credit decision and signature with the human underwriter. For lenders, the useful question is whether the software can work through their own loans under their own policies.

Copy the question before the acquisition

SAIGroup sits between software investing and software operation. Its stated approach emphasizes product investment, collaboration with management, and long-term growth. Customers encounter that strategy through separate businesses, with separate contracts and competing products. A hospital buyer and a supermarket buyer are shopping in different markets even when the ownership connects.

For a buyer, that creates a practical way to evaluate the offering. Ask the relevant portfolio company to demonstrate a specific job using representative data. A bank needs a defensible investigation trail. A clinician needs a recommendation that fits the encounter. A lender needs evidence attached to the file. These are different tests, and passing one says little about the others. The sensible comparison is with the work currently being done, including its delays and errors. A polished demonstration becomes useful when the people responsible for that work can inspect the output and decide whether it earns a place in their routine.

The lesson a smaller company can borrow is a question: where does useful knowledge already sit inside the business? Start there. Choose a task, understand its constraints, and judge the result in the workflow where somebody must use it. This is an inference from the portfolio, rather than a formula SAIGroup has proven universally.

Buying established software demands capital and integration skill. Applying AI also requires usable data, permission to use it, and people willing to trust the result. Without those conditions, an acquisition can leave you owning more complexity. The hospital television is a useful reminder: getting intelligence into everyday work takes a route to the user.