Homage Launches Esteem AI to Take the Paperwork Out of Home Care
The Singapore care company has turned the software it built for itself into a product for the sector, as an ageing population strains care teams. The agents prepare the work; people still approve it.
Homage, the Singapore-founded home care company, has launched Esteem AI, care management software for home care, aged care and disability support providers, and is opening it to the sector as a standalone business. The software runs a care agency’s operations in one place, from first enquiry to final invoice, and its AI agents handle the administrative work behind each visit: drafting visit notes from what a care worker recorded, suggesting a suitable worker for an open shift, following up on expiring certifications, and answering routine questions from families and staff. Nothing that affects a client’s care or a worker’s schedule happens without a person’s approval, and every action is logged. The product is available to providers in Singapore from today, with a global rollout to follow.
The launch is notable less for the technology than for where it comes from. Esteem AI is the system Homage built to run its own care services across Singapore, Malaysia and Australia, now packaged for others. It already has more than 50 care organisations on it, including the Home Nursing Foundation, and the company says more than 90 per cent use it as their everyday system. That is a different proposition from a vendor selling software into a sector it has never operated in, and it is also the claim a buyer should test.
What Homage launched
Esteem AI is described as agentic care management software. It replaces a care agency’s separate tools with one system for client intake, rostering, workforce management, incidents, customer support, billing, reporting and compliance, keeping each client’s history on a single record. On top of that, Homage is building AI agents to take on routine work before it reaches a care manager. When a visit ends, an agent drafts the note from the care worker’s record for the manager to edit or approve. When a shift is open, it suggests a worker based on skills, location and whether the client already knows them. It follows up on certifications about to expire, and answers routine questions from the record, passing anything needing judgement to a person with the context attached.
The design is deliberately conservative on autonomy. Nothing that affects care or scheduling happens without human approval, and every action is logged. Homage says more than 50 care organisations run on the software, over 90 per cent of them as their everyday care management system, and that in an average month more than 300 care managers rely on it. The Home Nursing Foundation, an independent charity that has provided nursing and medical care in Singaporean homes since 1976, runs its home personal care operations on it and now serves more than 1,000 care recipients a month.
Founded in Singapore
Homage was founded in Singapore in 2016 by Gillian Tee, a former management consultant who had also co-founded a New York travel start-up, and it has grown into one of the region’s better-known care platforms. It operates in Singapore, Malaysia and Australia, has a network the release puts at more than 30,000 Care Professionals, and says it has delivered more than a million hours of care. Its investors include Temasek’s Sheares Healthcare Group, East Ventures, Golden Gate Ventures and Jungle Ventures. The company has been building software for its own operations since it began, and it says it set out plans as early as 2020 for platforms other providers could use. Esteem AI is the result six years later.
The problem it addresses
The release frames the problem in terms care managers recognise. Rostering is still largely manual, and one cancelled shift can mean an evening of phone calls. Every visit note and certification has to be ready for an audit. Urgent matters, such as a missed visit or a change in a client’s condition, can get buried in group chats and inboxes. Much of this is managed across a rostering tool, a spreadsheet and a shared inbox that were never built to work together. Homage’s founder and chief executive, Gillian Tee, describes the surprise of discovering that everything around the care, the notes, the rosters, the forms and the calls to families, was as hard as finding carers in the first place.
Why provider-built matters
Software built by a care provider for its own use has a particular claim on credibility: it has been tested against real shifts, real audits and real families rather than a demo. It also carries a limitation. A system shaped around one provider’s way of working can require adjustment to fit another’s, and the release does not say how much configuration that involves. The distinction matters because the care sector is full of software that looks complete in a sales meeting and demands change management on the ground.
The demographic context
That problem is getting bigger. Singapore became a super-aged society this year: more than 21 per cent of citizens are 65 and above, up from 13.7 per cent a decade ago, and the Ministry of Health projects that one in four citizens will be 65 or older by 2030. The share is still climbing, because the largest cohorts have yet to cross the threshold. Every additional older person who needs care brings administrative work with them, and the workforce to do that work is not growing at the same rate. Esteem AI aims to close that gap, which helps explain why the company can reasonably describe demand as surging.
The workforce side of the equation is tightening too. Care work is labour-intensive, and providers across the region report difficulty recruiting and retaining staff, which is why the industry’s interest in automation is less about replacing people than about making scarce people go further. A tool that returns an hour a day to each care manager is, on the release’s logic, worth more than one that merely digitises a form.
The governance design is the interesting part
The most consequential design choice is where the automation stops. It would be easy, and commercially tempting, to let an agent write the visit note and file it, or assign a shift without a human. Homage has chosen not to. The system drafts notes for approval, proposes shift changes, and escalates judgement calls with context. It logs every action. In a sector where a mistake can affect someone’s health and where records must survive an audit, that is the right instinct, and it is also what distinguishes a tool a care manager will trust from one that threatens their accountability.
It is worth being clear that this is a design claim, not a verified outcome. The release does not say how much time the agents actually save, how accurate the drafted notes are, or how often managers accept a suggested worker. Those are the numbers that would show whether the agents reduce the administrative load or simply move it, and they are not in the announcement.
The category it enters
Agentic care management is already a competitive category, particularly in Australia and New Zealand, where home and community care software is mature. AlayaCare ships an agentic platform with agents for intake, scheduling, clinical and billing work, including a scheduling agent that can fill vacated visits on its own. Birdie sells rostering and care management to domiciliary care agencies. ShiftCare, Imploy, Lumary and others cover overlapping ground, several with AI features for note enhancement and shift matching. Esteem AI’s differentiation is less the agents, which rivals are also building, than its provenance: it is a product from a care provider that runs its own services, in a market, Singapore, where many of these vendors do not lead.
What buyers should ask
For a provider weighing it, the questions are specific. How much administrative time do the agents actually save, measured against a baseline? How accurate are the drafted notes, and how often do managers accept them? Moreover, how does the company secure the client record, where does it host the data, and what information does the AI share with model providers? How does the system handle uncertainty and escalate cases when necessary? What does the company charge, and how does it structure its pricing? And how portable is the record if the provider later leaves? It is also worth asking what happens if the model behind an agent changes, since care providers need continuity over years, not quarters.
What it signals
Esteem AI is a sensible move: Homage has monetised the software it needed, in a market where the need is real and growing, and it has done so with a conservative approach to autonomy that fits the stakes. The product is available now and has real users, which is more than most launches in this space can claim. The open questions are the ones that matter for adoption in care: measured time savings, note accuracy, and data handling. Homage bets that care providers will trust a system built by one of their own.
The evidence for that is the 50 organisations already using it, and the test is whether the agents deliver the time back that the company says they will. Whether that happens is measurable, and the company has not measured it publicly. That is the gap between a good product story and a proven one, and it is the gap a follow-up would seek to close.