Social housing providers operate as complex service organisations, managing numerous properties, serving diverse residents, and adhering to stringent regulations. If not effectively managed, this complexity can lead to inefficiencies, prolonged response times, risks for both residents and the organisation, and ultimately, resident dissatisfaction. Consequently, organisations should prioritise the relentless simplification of their processes. To achieve this, they must fully digitise their operations and implement systems that enable regular and simultaneous adjustments to their processes and systems to continuously enhance outcomes.
Digitalisation has long been a topic of discussion within the industry, rightfully so given its importance as a foundation for a learning and continuously improving organisation. However, this transformation has progressed slowly, and many organisations struggle to complete their digitalisation initiatives. Several organisations are hesitant to pursue digital transformation rapidly, often due to past errors resulting in wasted time, resources, and financial losses, or cautionary accounts of failure within the sector. This hesitancy leaves many organisations entrenched in the status quo, incurring substantial opportunity costs by delaying or avoiding the modernisation of their operations.
As the sector addresses the complexities of digital transformation, artificial intelligence (AI) is gaining mainstream prominence, intensifying the pressure on organisations to act decisively. While AI offers significant opportunities for organisations to optimise performance, it is not a panacea. Transitioning to an AI-powered enterprise necessitates an equivalent level of investment in process digitalization and system integration—areas where the industry has previously faced considerable challenges.
This is where Intelligent Automation (IA) comes in. IA denotes the strategic application of automation technologies—encompassing both traditional tools like Robotic Process Automation (RPA) and workflow engines, as well as contemporary AI-powered capabilities, such as those in large language models (LLMs)—to optimise processes, minimize manual effort, and empower systems to execute intelligent actions autonomously. It signifies a move from disparate automations to more cohesive, decision-making systems.
In this article, we explore how AI and automation—working together as part of an IA strategy—can address key challenges in social housing. We highlight the importance of digitalisation and offer practical guidance for moving beyond hype to real-world results. We make the case that while AI plays a critical role in enabling higher levels of intelligence and adaptability in automated systems, it is IA that provides the framework for meaningful and sustainable transformation.
“Forward-thinking organisations recognise the importance of a structured, high-quality data foundation and a well-architected software infrastructure as critical foundations without which AI cannot function reliably, reach its full potential, be trusted to support operational decision-making or deliver consistent outcomes at scale.”
The Promise and Limits of AI for Complex Service Organisations
Recent regulatory action by the Regulator of Social Housing highlights the urgency for modernising processes and data management. Over the past two years, the most common reasons for intervention have included health and safety compliance failures, fire safety breaches, inadequate management of damp and mould, widespread issues in repairs and maintenance services, and shortcomings in data management and record-keeping. In numerous instances, housing providers were unable to provide up-to-date information about property conditions, safety certificates, or outstanding works, thereby exposing both residents and organisations to significant risk. These examples highlight that the progression towards AI and automation must be coupled with solid digital foundations and comprehensive visibility across compliance and operational data.
Accordingly, for complex service organisations, the AI revolution represents both an opportunity and an imperative. AI offers significant potential for social housing providers to transition from reactive service models to more proactive and predictive operations. It can assist organisations in making more informed decisions, automate tasks that weren’t possible before, and deliver enhanced outcomes for residents and staff alike. The real challenge for organisations isn’t simply deciding whether to adopt AI—it’s how swiftly and effectively they can build the essential data and software foundations that make AI truly valuable.
Nonetheless, AI is not a quick fix for every problem. It works best when processes are already clear, data is reliable, and workflows are digitised. If records and operational data are scattered across PDFs, spreadsheets, and emails, it becomes problematic for AI models to produce precise and consistent insights. One might be inclined to believe that all unstructured data—emails, reports, and raw logs—can simply be accumulated in a data lake for AI agents to interpret subsequently. Although AI has made remarkable advancements in processing unstructured information, depending on this “data swamp” method remains risky. AI yields the best results when it has access to well-structured, contextualised data, rather than having to discern meaning from a disorganised assortment of documents.
Furthermore, AI’s help shouldn’t be confined to interpreting static data, summarising documents, or suggesting actions for users to execute manually. If the objective is to reach intelligent automation (IA), the next best actions should ideally be undertaken by the systems automatically, which necessitates systems capable of programmatic action. In this context, AI can function as a force multiplier for existing automation tools, facilitating solutions to potential automation challenges that were previously impossible to address with pre-AI technologies.
AI Readiness: Building the Foundations for Intelligent Automation
Successful deployment of AI depends on high-quality, well-organised data and mature digital processes. Many organisations fail to grasp just how fundamental this foundation is. Prior to considering AI adoption, leaders should ensure that workflows are digitised, data is accessible and reliable, and core business processes are standardised across teams and systems.
A common obstacle in social housing is fragmented data—information scattered across PDFs, spreadsheets, emails, or legacy databases. This fragmentation creates ‘data swamps’ that undermine both automation and AI initiatives. To achieve true AI readiness, organisations must dedicate resources to cleaning and structuring data, migrating legacy records into modern platforms, and adopting standard data schemas that support interoperability. The technology stack must be able to integrate modern automation tools and AI models with existing business systems. Investing in open APIs, integration platforms, and modular software enables flexibility and scalability—essential for any intelligent automation strategy.
Robust data governance is crucial. This encompasses establishing clear data ownership, stringent data quality controls, and procedures for routine auditing and maintenance. Effective governance also supports regulatory compliance and cultivates trust with staff and residents.
Lastly, establishing AI readiness extends beyond mere technical work. Teams must foster a culture of digital adoption and continuous improvement, which includes preparing staff for change. Most employees worry about the risk of job loss due to automation and AI, so a progressive organisation should invest in training programmes and transparent dialogue to help staff understand the benefits, address concerns, and build confidence. This minimises opposition and unlocks innovation at every level.
"A common risk is underestimating the importance of a structured software and data foundation. Without clean, connected, and well-architected systems, AI struggles to deliver insights that are timely, reliable, or actionable. Without that foundation, even the most advanced models can fall short of their potential and undermine trust in AI-supported decision-making."
When AI is (and Isn’t) the Right Tool: The Full Automation Toolkit
AI constitutes merely one component of a comprehensive automation toolkit that encompasses RPA, Machine Learning (ML), Natural Language Processing (NLP), rule-based workflow automation, Business Process Management (BPM) platforms, and process mining tools. Increasingly, this toolkit includes integration Platform-as-a-Service (iPaaS), no-code, and low-code platforms. Decision automation engines and orchestration platforms further empower organisations to automate end-to-end workflows, coordinate multiple technologies, and continuously optimise operations whilst empowering business users to adapt processes without requiring deep technical expertise.
The efficacy of each tool fundamentally depends on the specific problem being tackled. Not every challenge requires generative AI—established methods often prove more reliable whilst introducing less risk and complexity. Where consistent, repeatable outputs are essential, organisations should favour conventional deterministic approaches such as rules-based automation, machine learning, optimisation algorithms, or conventional software logic, rather than relying on non-deterministic AI models like LLMs that may produce variable results for identical inputs.
RPA excels at automating tasks within legacy systems lacking APIs by mimicking human interactions and bridging disconnected systems. Traditional workflow automation and integration tools prove invaluable for streamlining routine processes, reducing errors, and saving time—often without AI's associated overhead.
For advanced analytics and predictions, ML typically represents the optimal choice, particularly for structured data and clearly defined prediction tasks. ML models can forecast maintenance requirements, classify risks, or detect anomalies with greater transparency and control than LLMs. These pre-LLM approaches prove preferable because they are purpose-built, easier to validate, and typically more efficient in production environments.
LLMs excel at handling unstructured data, extracting insights from free text, and enabling natural language interactions. They deliver value where automation involves communication, document analysis, or requires contextual reasoning beyond traditional tools' capabilities. However, LLMs can bring added complexity, require more oversight, need continuous calibration, and may not deliver the same level of reliability, cost-efficiency, or interpretability as conventional approaches for routine, well-specified tasks.
Therefore, each use case merits individual assessment to determine the most effective tool. Whilst AI can enhance automation intelligence and adaptability, many scenarios benefit from proven automation methods or classical ML that deliver superior results with reduced complexity. The ultimate objective should be intelligent automation—leveraging the optimal combination of tools, including but not limited to AI, to achieve streamlined, effective, and sustainable outcomes.
Implement Responsible AI
As organisations increasingly embed AI into their operations, they encounter new data privacy and ethical challenges, and often find it difficult to clearly define and manage the associated risks. If organisations do not give sufficient attention to responsible AI practices, they may encounter challenges such as biased or unfair results, loss of trust, reputational damage, or even regulatory scrutiny. A lack of transparency can make it hard to explain or justify AI-driven outcomes, while inadequate moderation, auditability, or ongoing calibration may allow issues to persist or go unnoticed. Setting clear boundaries for how and where AI is used, maintaining transparency with staff and residents, and routinely auditing and refining systems can help organisations reduce these risks and foster confidence as they adopt more advanced technologies.
"AI systems must be transparent, fair, and subject to regular audits to guard against bias and discrimination. To achieve this, it's essential to design AI solutions that are inherently auditable—particularly those employing self-learning models that evolve over time."
Asset Upkeep and Compliance: Practical Use Cases for AI and Automation
In social housing, operational areas such as responsive repairs, planned preventative maintenance, asset management, and compliance are especially well-suited for the application of AI and automation. These domains are rich with complex, repetitive, and data-driven tasks—creating significant opportunities to achieve efficiency gains, enhance service quality, and strengthen risk management through technology. Key use cases include:
Responsive Repairs: Automation can streamline the logging, triaging, and assignment of responsive repairs. AI can help classify and prioritise requests based on urgency, asset criticality, or resident vulnerability. Intelligent systems can also analyse historical data to spot recurring issues or predict when assets are likely to fail.
Planned Preventative Maintenance: While automation tools can already manage long-term schedules for cyclical maintenance tasks, AI models can optimise these schedules based on asset condition, usage patterns, and available resources, reducing costs and minimising service disruption.
Predictive Maintenance: Machine learning algorithms can analyse data collected from IoT devices and sensors, along with maintenance logs and environmental factors, to forecast asset failures before they happen. By leveraging real-time sensor information from equipment and building systems, providers can intervene proactively, reducing emergency repairs and extending asset life.
Planned Capital Works: Optimisation algorithms can plan capital projects, integrating data on asset condition, compliance requirements, and budget availability. AI can be used to prioritise projects based on risk, budget, and potential impact, making capital planning more strategic and data-driven.
Compliance (Building Safety and Regulation): Compliance operations are among the most complex in housing and require investment in intelligent systems and end-to-end digitalisation. Without this, organisations remain reliant on manual workarounds and scattered documents, making compliance management difficult. Intelligent automation supports compliance by tracking obligations, sending reminders, and ensuring records are audit-ready. AI can help further by extracting and structuring historical data from PDFs and spreadsheets, and by processing unstructured or non-digital certificates.
No-access Risk Prediction: AI models can use historical patterns and resident data to predict which appointments are likely to result in "no access," allowing for proactive communication, better scheduling, and fewer failed visits.
Contractor Performance Analysis: Automation can collate data from work orders, feedback forms, and inspection results. Intelligent decision systems—whether or not they use AI—can optimise the distribution of work orders across the supply chain, using real-time data on contractor performance, availability, and location, provided the organisation has systems capable of storing and accessing this data. AI can then assess contractor performance over time, identifying trends and supporting evidence-based procurement and contract management decisions.
Data Integration and Analytics: Many providers use multiple software solutions for housing management, CRM, asset management, and compliance. AI and automation can act as an intelligence layer across these systems, surfacing insights, identifying inefficiencies, and enabling more informed decision-making by presenting a unified view of asset health, compliance status, and operational risk. Also, when organisations fully digitalise their operations massive amount of transactional data gets generated. While this information can be very useful, making sense of vast amounts of data using BI tools alone can be challenging. AI can really help in making sense of historical and trend data, identify patterns, suggest ideas to improve the KPIs and regularly send these insights to the right people in the organisation when automated.
In all these areas, the combination of automation and AI delivers tangible benefits: reducing manual effort, improving service reliability, supporting proactive asset management, and enhancing resident satisfaction. As digital maturity increases, housing providers can continue to unlock more value from these tools—moving from reactive to predictive and strategic asset management.
From Hype to Action: Realistic Next Steps for Housing Providers
As the conversation around AI continues to grow, it is easy for housing providers to become distracted by promises of transformational change or wait for the next big technological breakthrough. However, meaningful progress is more often achieved through a balanced approach that combines ambition with practical action. Rather than viewing AI as a silver bullet, organisations can focus on intelligent automation (IA) as the primary goal—leveraging AI where it is most effective, but never losing sight of the fundamentals of process simplification and digitalisation. In some cases, traditional software logic or process redesign may deliver better results with less complexity. It’s important to assess each use case on its own merits, recognising that IA is the overarching goal—achieved through a combination of traditional automation tools, AI-enhanced automations, and AI agents—while choosing the most effective technologies to support that goal. By doing so, housing providers can deliver real value, reduce risk, and set the foundation for ongoing innovation.
To move forward effectively:
Remember that continuous simplification and digitisation of processes should remain a core focus, whether pursuing IA with or without AI.
Set clear, practical goals that align with organisational priorities.
Identify processes likely to benefit most from automation or AI—typically those that are repetitive, data-driven, or critical to service delivery.
Leverage automation and AI capabilities already available in existing software platforms to minimise cost and implementation risk.
Pilot new tools in small, controlled projects, gather feedback, and refine approaches before scaling successful solutions across the organisation.
Avoid treating IA or AI as large, sequential “big bang” projects that require months or years of planning before delivering value. Instead, pursue incremental improvements, adapt quickly, and stay focused on solving real business problems.
Maintain an iterative, problem-focused approach to ensure technology delivers measurable value to staff and residents, rather than becoming a distraction or source of unnecessary complexity.
“IA (intelligent automation) is the destination: it unites proven automation and the latest AI to deliver real business value. AI accelerates progress, but IA is the unifying framework for lasting impact.”
Conclusion: Automation for Impact, AI for Advantage
The future of social housing will be shaped by organisations that master simplification, digitalisation, and intelligent automation. Those who invest in strong digital foundations, adopt a pragmatic approach to automation, and use AI to augment—not replace—sound operational practice will unlock new levels of efficiency, resilience, and service quality. By focusing on IA as the goal, and leveraging AI where it adds the most value, housing providers can deliver both immediate improvements and long-term strategic advantage.
Emre Kazan, Co-founder and CPTO at Plentific
