Agentic AI is coming to the housing sector. Trevor Hampton, director of housing solutions at NEC Software Solutions, explores the challenges and opportunities for housing providers to get the maximum value from this emerging technology.
The publication of Alan Turing’s groundbreaking ‘Computing Machinery and Intelligence’ paper in 1950 was arguably the moment when the question ‘can machines think?’ first entered public consciousness.
Since then, artificial intelligence (AI) has come a long way. Many of us now use it to enhance our photos, automatically filter spam from our inboxes and get instant action points from meetings.
This generation of AI has also made its way into the housing sector, where housing providers are increasingly using it to automate time-consuming tasks such as managing repairs and rental payments or to predict when heating systems need to be upgraded.
Housing providers that have adopted AI have used it to save time, cut costs and improve tenant satisfaction. And the next generation of AI is already here, in the form of ‘agentic AI’.
What agentic AI is and isn’t
Traditional AI has its limits; agentic AI is designed to take the technology a step further.
A housing provider might want to understand the best way to reduce the risk of damp and mould in households with young families over the next 12 months. There are significant differences between how traditional AI and agentic AI could help with this.
Traditional AI is designed to carry out well-defined requests, typically answering a single question or applying a fixed set of rules. In this example, it could identify which maintenance requests are most urgent or flag which households meet certain criteria.
The technology has some capacity to mimic human decision-making but its outputs are constrained by the specific instructions it is given. It also lacks the ability to work through a series of related actions or adapt over time to deliver an end-to-end solution to a problem.
In contrast, agentic AI does far more than follow a set of instructions. Given a challenge or end goal, it can access multiple data sources and then reason, plan and adapt to deliver a practical solution.
With the task of reducing damp and mould in the most vulnerable households, agentic AI would analyse historical repairs data direct from the housing provider’s HMS and other core systems, identify properties with recurring moisture problems, cross-check this with occupancy levels and vulnerability indicators and forecast which homes are most at risk and at what times of the year.
The technology could then instantly analyse all of this information to suggest a proactive plan, which might include a list of properties needing ventilation checks, in priority order according to the vulnerability of the residents, along with recommendations for maintenance work or possible improvements to the properties’ insulation. The AI agent could even draft a series of personalised communications to the residents affected, with branding and messaging direct from the housing provider’s own guidelines, then coordinate contractors’ availability to carry out the work and book follow-up visits for all relevant housing officers.
This represents a fundamental shift in the way housing services can be managed and delivered in the future. Housing providers that start thinking about how these AI agents could support them will be best placed to deliver real improvements for residents and staff alike.
Agentic AI will continue to mature over the next three to five years. To take full advantage of the shift, there are three strategies housing providers should start with now to benefit from this evolution.
1. Get data-ready
Agentic AI can analyse structured datasets, such as property records or maintenance logs, alongside more unstructured information, such as meeting notes, observations from housing officers following a visit and residents’ feedback forms. Giving AI agents access to a broader range of information can reduce the risk of bias and uncover subtle insights that analysing structured data alone might miss.
Data should be integrated, accurate and from a trusted source to take full advantage of agentic AI for semi-autonomous and autonomous decision-making. The technology can then link all relevant details in real time to anticipate problems, prioritise interventions and free staff to focus on delivering safe and comfortable homes for residents.
2. Build skills across your organisation
Preparing housing officers and staff for agentic AI is as important as preparing the data. Upskilling staff early helps them to recognise where AI agents can add value and which use-cases will have the greatest impact on residents and the organisation.
With the right training, housing staff can also gain confidence and a deeper understanding of how the technology interacts with their existing legacy systems so they know how to use it effectively. This could include a combination of training programmes, cross-departmental workshops and a team of agentic AI champions who can build trust in the technology and create a culture ready to take full advantage of agentic AI.
3. Choose the right technology partner
To fully benefit from agentic AI, new systems must be intuitive and easy for people to navigate. But the real value lies in what these innovative solutions can deliver for the housing sector.
A housing officer should be able to ask direct questions such as, “what are the key problems in my repairs today?” or “which complaints require urgent attention?” and get the answers they need instantly. This eliminates the time-consuming task of searching for information manually and frees staff to focus on keeping homes safe and residents supported.
The right technology partner will also ensure housing providers receive the guidance needed to adopt AI agents effectively and align them with their individual business goals, whether that’s to expand their assets, introduce smart technology into homes or predict and prevent households from falling into arrears.
Trevor Hampton is the director of housing solutions at NEC Software Solutions.
