Key Stats at a Glance
Oxford University, one of the world’s leading higher education institutions, sought to explore how AI could transform its Professional Services. Administrative functions — from HR to finance and student support — faced increasing demands and rising costs, while academic and research excellence remained the institution’s primary focus.
The University recognised that AI had the potential to reduce administrative burden, unlock efficiency, and enhance staff and student experiences. However, a roadmap was needed to move from theory to practical adoption.
Hudson & Hayes was engaged to develop an AI roadmap, building literacy, identifying opportunities, and charting a path to scalable adoption.
Key issues included:
Hudson & Hayes applied its GenAscend methodology, tailoring the approach to a university environment.
Hudson & Hayes helped Oxford University move from curiosity to clarity on AI adoption in Professional Services. By embedding literacy, defining a prioritised opportunity pipeline, and creating a roadmap, the University is now positioned to leverage AI in a strategic, scalable, and responsible way.
This foundation ensures Oxford can continue to focus on academic and research excellence, supported by professional services that are efficient, future-ready, and digitally enabled.
When a UK transport organisation introduced Microsoft Copilot in procurement, adoption quickly became the biggest hurdle. The proof of concept showed real potential, but most employees weren’t engaging with the tool — meaning its AI productivity benefits were being left on the table. In this discussion, Chelsea and Gareth explore how tailored training, bite-sized learning, and Copilot champions helped turn the rollout into a model for successful digital transformation.
Chelsea: Gareth, can you set the scene for us? What challenge was this UK transport organisation facing with their Copilot rollout?
Gareth: The big issue was adoption. They had launched a proof of concept across procurement, but engagement was patchy. Some staff were experimenting, but most weren’t using Copilot consistently. That meant the tool’s potential to actually reduce admin and free up time wasn’t being realised.
Chelsea: So how did we approach fixing that?
Gareth: We co-created a tailored curriculum with the client. Rather than generic training, we focused on real employee pain points. We also introduced Copilot champions who could support their colleagues directly, which made it feel much more relevant.
Chelsea: What did the training look like in practice?
Gareth: We designed 15-minute, bite-sized sessions. The idea was to spark curiosity and show “the art of the possible” without overwhelming people. That format kept energy high and made it easy for staff to fit into their schedules.
Chelsea: And what kind of impact did this approach deliver?
Gareth: Over 250 procurement staff across all regions took part. Crucially, their feedback shaped the content as we went, so training stayed practical and impactful. By the end, Copilot wasn’t just a tool, it became a real productivity driver for the team.
The results were significant: over 250 procurement staff across all regions adopted Microsoft Copilot, embedding it into their daily workflows to save time, reduce admin, and improve efficiency. By focusing on practical use cases and AI literacy, the organisation turned a slow start into a model for successful Copilot adoption.
This case highlights how businesses can unlock the full value of AI in procurement and drive measurable digital transformation by combining technology with people-focused change.
For more than a decade, Robotic Process Automation (RPA) has been the backbone of back-office efficiency. By mimicking keystrokes and clicks, RPA freed employees from repetitive, rule-based tasks like invoice entry and claims processing. It delivered measurable cost savings and faster processing times. But RPA has limitations: it breaks when processes change, struggles with unstructured data, and ultimately only scratches the surface of what’s possible with automation.
Today, a new wave is taking shape. AI Agents and Agentic AI are redefining how organisations think about digital transformation. The shift is not just from faster scripts to smarter bots — it’s from automation as task execution to automation as orchestration.
This journey can be mapped as an automation maturity curve with three stages:
Organisations that understand and climb this curve will move from incremental savings to step-change strategic value.
RPA thrives in high-volume, structured processes. It is best suited for tasks that follow clear rules and rarely change. Examples include:
The value case for RPA is straightforward: efficiency and cost reduction. By removing repetitive keystrokes, organisations gained speed and accuracy. However, fragility is its main weakness. A small change in process or system layout can break an RPA bot. And because RPA relies on structured inputs, it cannot handle the unstructured data that dominates knowledge work — such as emails, free-text fields, or conversations.
RPA is therefore the entry-level stage of automation maturity. It is useful, it delivers savings, but it is not transformative.
Where RPA mimics clicks, AI Agents understand context. Powered by large language models (LLMs) and integrated with enterprise tools and data, AI Agents can:
The breakthrough is in handling unstructured data. Emails, documents, and chat logs can be understood and acted upon. Unlike RPA, AI Agents are not brittle — they adapt to new inputs, guided by human oversight.
This makes AI Agents a bridge between efficiency and intelligence. They do not just automate tasks, they augment knowledge work. For employees, this means less time searching for answers or updating systems, and more time solving higher-value problems. For customers, it means faster, more personalised service.
As an example, a service desk agent supported by an AI Agent can resolve routine tickets instantly while escalating complex cases to human experts. The result is not only efficiency but a better end-user experience.
The frontier of automation is Agentic AI. Instead of following a single-task instruction, Agentic AI systems can:
Imagine the goal “optimise the procurement cycle.”
This is not just automation, it is orchestration. A network of AI agents collaborates to achieve an outcome, with minimal human input. Humans set direction, provide oversight, and make final calls.
The business value is transformational. Instead of speeding up existing processes, Agentic AI enables organisations to reimagine how work is organised. Procurement, supply chain, clinical scheduling, or customer onboarding can shift from sequential, human-driven tasks to parallel, AI-driven ecosystems.
The three stages form a maturity curve:
Importantly, each stage does not replace the last. They build on one another. RPA still has its place for structured processes. AI Agents elevate knowledge work. Agentic AI unlocks orchestration and adaptive decision-making.
Leaders must assess where they are today and design a roadmap for progression. A balanced portfolio will combine all three, applied to the right contexts.
Why does this matter now?
For organisations, the message is clear. Automation is no longer just a tool for cutting costs. It is a strategic lever for redesigning operations.
The companies that win will not be the ones who “just add AI.”
They will be the ones who climb the automation maturity curve, using the right tool for the right context:
We are moving from automation as cost-cutting to automation as strategy. The question is no longer “What can we automate?”
It is “How do we design our operating model for a world of autonomous, multi-agent systems?”
Time and time again, we hear from clients that they’ve rolled out Microsoft 365 Co-Pilot licences across their organisation, but they’re unsure whether they’re getting the benefit they expected.
In a large business, this isn’t a trivial investment. At £30 per user per month, rolling it out to a team of 1,000 employees means you're spending £30,000 per month — £360,000 a year.
It’s a serious commitment.
The potential, however, is real. Studies suggest you can unlock 2–4 hours of time saved per person per week — but only if people actually know how to use it, and more importantly, how to apply it within the flow of their work.
I've personally leaned into the use of AI in my daily work — but I know that doesn’t represent everyone. People are at different stages of the change curve. Some are excited, some cautious, some overwhelmed. Your rollout must reflect that. Not everyone needs the same training or use cases. Segment your audience and meet them where they are.
Don’t assume people will work out how to use Co-Pilot on their own. Curate a set of standard use cases for each function or role, alongside prompts that get results. Go further by identifying which use cases should be leveraged department-wide, and build these directly into workflows using Co-Pilot Studio.
Co-Pilot is only as good as the data it has access to. Audit where your information is stored. Are permissions right? Are Teams and SharePoint sites structured logically? Is your knowledge base accessible and up to date? Poor data = poor answers.
Prompting is a skill. It’s not just about asking the right question, but understanding prompt chaining — how to iterate, build on answers, and think in systems. Most people don’t need a feature tour; they need a mindset shift in how to work with a digital assistant.
Create light governance that enables teams to develop, share, and re-use effective Co-Pilot use cases. Track what’s working. Encourage teams to propose their own prompts and use cases, then socialise the ones that can scale across the business.
This is key: take a process-centric approach, not a tool-centric one. Map where Co-Pilot fits into your current workflows, from preparing for a client meeting to summarising Teams threads or building a project plan. Co-Pilot isn’t just for shortcuts — it can fundamentally reshape how work flows.
Co-Pilot isn’t magic, and it’s not always right. Set realistic expectations early — it's a productivity partner, not a replacement for critical thinking. Use bite-sized training to reinforce this message, and highlight both its strengths and limitations.
Make Co-Pilot part of onboarding, team meetings, and even personal development plans. Encourage teams to challenge each other: “Could we have done this faster or better using Co-Pilot?” This embeds a culture of continuous improvement — and AI adoption becomes second nature.
Identify Co-Pilot Champions in each department who can lead by example, support others, and share what’s working. One organisation we came across ran a brilliant “Prompt of the Week” campaign. It created a buzz and got people experimenting. Recognise and celebrate those who lean in.
Build a comms strategy that regularly shares success stories, new prompt ideas, and usage stats. Highlight the time saved, the creative breakthroughs, and the tasks Co-Pilot is now handling. Storytelling is your biggest lever for cultural change.
The licence fee is fixed — but the value you get from it isn’t. Co-Pilot can fundamentally change how your teams work, collaborate, and think. But that only happens with the right mix of enablement, structure, and culture.
Don't just deploy it. Operationalise it.
If you’d like help mapping out high-value use cases, training your teams, or embedding Co-Pilot into your workflows, we’d love to talk.
Introduction: In this artificial intelligence in procurement case study, we highlight how a global organisation transformed its procurement processes, saving over 10,000 hours annually. By implementing AI-powered solutions, Robotic Process Automation (RPA), and advanced analytics, the organisation automated manual tasks, gained valuable insights, and significantly improved procurement efficiency. We also focused on educating the team on the use of Machine Learning and Predictive Analytics in procurement, which enhanced decision-making and forecasting accuracy.
Challenges: The organisation encountered several challenges that hampered procurement efficiency, such as manual data entry, outdated forecasting methods, and limited visibility into supplier performance. These issues caused delays and prevented the procurement team from concentrating on more strategic initiatives.
Approach: We addressed these challenges by implementing AI and automation technologies while providing tailored training on modern procurement strategies like Machine Learning and Predictive Analytics.
Results: Our initiatives led to a total time savings of 10,000 hours per year. Key results included:
Conclusion: This artificial intelligence in procurement case study showcases the transformative power of AI and automation in procurement. By automating key processes, educating teams on advanced procurement technologies, and optimising existing systems like SAP Ariba, we achieved over 10,000 hours in time savings, allowing the procurement team to focus on higher-value strategic initiatives.
Fill out the form, and we’ll send you a copy! It’s a comprehensive guide to implementing AI in procurement, complete with practical use cases, expert insights, and strategies to help you streamline processes and drive efficiency in your organisation.
The digital transformation in Human Resources is reshaping the way businesses operate, moving towards a self-service model that aims to streamline operations and empower employees. This evolution brings a mix of reactions as it challenges the traditional roles and functions within HR departments.
The shift towards self-service technology in HR mirrors the convenience and user-centric design of modern digital experiences. By enabling employees to manage their own HR tasks, companies can reallocate their HR resources to focus on more strategic, impactful work. Balancing the efficiency of technology with the need for personal support is crucial to maintaining the human element that is core to HR.
The integration of self-service options with automation is revolutionising HR processes, enhancing both speed and user experience. Advanced HR/ERP platforms like Workday and SAP SuccessFactors are becoming even more efficient with the addition of automation, pushing the boundaries of what's possible in HR operations.
Generative AI (GenAI) is at the forefront of the HR transformation, promising to significantly improve the entire employee experience. With the potential to automate up to 30% of HR tasks, GenAI not only streamlines operations but also opens up new opportunities for deeper employee engagement.
Opportunity: AI-driven digital assistants can provide 24/7 support for HR policy inquiries, streamlining information delivery and freeing HR professionals to focus on more complex, strategic tasks.
Opportunity: AI tools can analyse and summarise employee feedback, offering insights to refine employer branding and align it more closely with employee expectations.
Opportunity: AI can automate the creation of job descriptions and personalise the onboarding experience, enhancing the efficiency and effectiveness of the talent acquisition process.
Opportunity: Routine HR tasks, such as processing leave requests, can be automated to reduce administrative workload and improve overall efficiency.
Opportunity: Automation can facilitate compensation benchmarking and analysis, ensuring that compensation packages are competitive and equitable.
Opportunity: The development of e-learning content can be automated to create engaging and personalised learning experiences for employees.
Opportunity: AI can tailor career development paths to individual employees, supporting personal growth and job satisfaction.
Opportunity: AI can efficiently process and analyse employee survey data, providing valuable insights into engagement and satisfaction levels.
Opportunity: Automating the offboarding process ensures a smooth transition for departing employees and maintains the integrity of organisational systems and data.
The potential of automation and AI in HR is vast, particularly with GenAI, which offers unprecedented personalisation and efficiency. This strategic evolution marks the beginning of a new era in HR, where technology not only streamlines operations but also significantly enhances the employee experience. As HR embraces these advancements, the focus shifts from administrative duties to strategic initiatives that nurture a dynamic and engaging workplace culture.
In the world of Robotic Process Automation (RPA) and Intelligent Automation (IA), success stories often overshadow the complexities and challenges encountered along the way. Drawing on extensive experience across numerous RPA and IA projects, this blog aims to shed light on the pitfalls that can undermine your business case and how to navigate them effectively.
Complex system landscapes and siloed technologies can significantly inflate development and maintenance costs. A case in point involved a global client with a high-volume, time-consuming process ripe for automation. The catch? Each region operated a different instance of SAP, necessitating multiple bots. Understand IT Architecture: A deep dive into the IT landscape is essential for crafting a scalable automation strategy that accounts for regional variations.
The disconnect between automation initiatives and the broader technology roadmap can impact the payback period, especially with impending platform replacements. Integration with Technology: Integration with the technology roadmap ensures that automation efforts complement upcoming system upgrades or ERP implementations, safeguarding your investment.
Often, there's a gap between documented processes and their real-world execution, leading to underestimations of process complexity. Understand Process Nuances: Engage directly with process participants and/or leverage process mining tools to capture the full scope and variability, ensuring a realistic assessment and preparation for automation.
The allure of automating low-cost or outsourced functions must be balanced against the total cost of ownership of RPA or IA solutions. Financial Prudence: Conduct thorough cost-benefit analyses to ascertain the financial viability of automation projects, particularly when considering the replacement of low-cost human resources.
Even the smoothest projects can falter in the adoption phase due to inadequate change management. Effective Communication: Effective communication, education, and the appointment of change champions are critical to ensuring user acceptance and maximising the utilisation of new automation solutions.
By sharing these insights, the aim is to help organisations maximise the benefits of their automation deployments, ensuring a strategic, informed approach that navigates the complexities of RPA and IA projects.
The launch of Chat GPT by OpenAI has created excitement in many organisations, offering glimpses into the vast possibilities of chatbots for answering common questions. It's tempting to assume that one can simply direct a Large Language Model (LLM) to a comprehensive knowledge base. However, this approach can risk losing control over responses, especially when questions overlap across multiple datasets.
Here, we explore different chatbot architectural options, highlighting their nuances, and their associated advantages and disadvantages.
These platforms involve defining intents, entities, and utterances to develop conversational flows.
Pros:
Cons:
This involves an amalgamation of chatbot's understanding with LLM's expansive knowledge. For instance, upon receiving a query, the chatbot platform may forward complex queries to the LLM.
Pros:
Cons:
Figure 1: Example architecture | LLM integrated with Amazon Lex
Directly interfaces the LLM with a customised user interface, bypassing intermediary platforms.
Pros:
Cons:
Open Source Models: Although OpenAI often emerges as a top choice, the world of LLMs is vast and evolving. There's a surge in open-source LLMs. Models such as BAAI’s Aquila, EleutherAI’s GPT-J, Google’s Flamingo, and TII’s Falcon LLM are some notable names from a growing list. The open-source ethos is alluring, especially with platforms like Hugging Face serving as repositories. Such platforms often provide fine-tuned versions of foundational LLMs.
Addressing Data Privacy: When using an LLM, one might be concerned about data security. OpenAI's recent Azure service emphasises data privacy, assuring that data is not used to refine models, stored only for 30 days, and remains isolated from third-party access. For detailed insights, you might want to visit OpenAI's data privacy page.
Choosing the right chatbot architecture hinges on an organisation's specific needs. For those already equipped with chatbots, a full shift can be cumbersome and risky. However, incorporating an LLM can notably elevate response quality and expand its capabilities. The allure of open-source models is undeniable, yet understanding the implications for data privacy remains vital. The key lies in striking a balance between leveraging innovation and maintaining practical control.
Automation / Artificial Intelligence is reshaping the way organisations operate, driving efficiency and freeing up valuable time. But the question remains - how best can we utilise these newfound productivity gains? Here are some possibilities to consider.
Automation often results in small time savings across many roles rather than substantial savings for a few. In these scenarios, think about consolidating these savings for broader benefits - this could mean new training opportunities, team collaboration initiatives, or collective process improvements.
The ideal choice depends on several factors:
Don't forget to consider the wider economic climate and your organisation's performance. For example, if your business is growing with new roles opening up, you could opt to reallocate tasks to existing staff rather than recruiting new team members.
Any option you choose will bring change, and change can be unsettling. Keep your team in the loop with clear, consistent communication. Develop a robust change management strategy to guide your organisation through the transition, involving HR where necessary - especially if job roles are affected.
Choosing how to use productivity gains from automation is a careful balancing act. By considering your options in depth, understanding your business context, and following a systematic process, you can arrive at a decision that best serves your organisation and its people.
Conversational AI, a technology initially focused on external customer-facing processes, is now transforming back-office operations. The rise of generative AI has expanded the range of use cases, offering significant potential to automate repetitive tasks, create additional channels for information retrieval, and enhance the internal customer experience. Procurement teams often spend considerable time handling enquiries from internal stakeholders, many of which could be resolved independently. As a result, introducing conversational AI and chatbot technology can lead to substantial time savings.
A study by The Hackett Group discovered that procurement personnel dedicate up to 60% of their time to tactical activities, such as addressing routine questions from the business, which could be automated through Conversational AI (The Hackett Group, n.d.). This presents a tremendous opportunity for organizations to achieve increased efficiency and productivity by implementing Conversational AI in procurement processes.
It is important to note that the terms conversational AI and chatbots are frequently used interchangeably, but they do not mean exactly the same thing. Conversational AI encompasses the wider domain of artificial intelligence that allows machines to comprehend and respond to human language. Chatbots, in contrast, are a specific application of conversational AI designed to interact with users through natural language formats, typically via text or voice-based interfaces. Chatbots employ natural language processing (NLP) and machine learning (ML) algorithms to understand user intent and respond in a manner that simulates human conversation. Although chatbots represent just one facet of conversational AI, their increasing popularity across various industries, including procurement, can be attributed to their ability to automate routine tasks and deliver a more streamlined user experience.
It's important to understand basic ChatBot architecture. By understanding basics about how a ChatBot responds to user queries it can bridge the gap between business and technology and spark ideas on potential use cases.
The chatbot architecture works as follows:
The reach of the chatbot depends on the number of intents it can understand and respond to accurately. The more intents a chatbot can handle, the greater its reach. Similarly, the more entities a chatbot can extract, the more personalised and effective its responses will be.
Chatbot responses can be categorised into three main types: hard-coded responses, data-driven responses, and generative responses. Each type has its advantages and limitations, depending on the specific use case and requirements.
Hard-coded (text-based, radio buttons, links):
Chatbots with a natural language understanding (NLU) engine use hard-coded responses like text, radio buttons, or links for predetermined answers to specific user inputs. The NLU engine processes user inputs, allowing the chatbot to comprehend the conversation's context. The chatbot selects a hard-coded response based on the identified intent, providing a structured and controlled conversational flow. However, this approach lacks the flexibility of advanced, generative models.
Data-driven:
Data-driven chatbots retrieve information from back-end systems like databases or APIs. They often combine rule-based or generative techniques with data retrieval, providing users with accurate, up-to-date information. Data-driven chatbots are suited for tasks requiring specific, dynamic data.
Generative:
Generative chatbots, like GPT-4, use machine learning algorithms based on natural language processing (NLP) and natural language generation (NLG) techniques. They generate responses by predicting appropriate word sequences based on user input, enabling more diverse and contextually relevant replies.
Conclusion: Understanding the different types of chatbot responses is essential when selecting the best approach for your specific use case. While hard-coded responses provide a more structured and controlled conversational flow, data-driven responses offer dynamic information retrieval, and generative responses enable a more flexible and contextually relevant conversation.
Procurement Policy & Process
Order Management
Supplier Relationship Management
Sourcing
Spend and Budget Tracking
Guided Buying & Purchasing
Supplier Collaboration & Self Serve
Description: A chatbot platform like Amazon Lex or Dialogflow offers pre-built solutions that can be customised to your organisation's needs. These platforms provide a balance between flexibility and convenience and include built-in NLU (Natural Language Understanding) and NLG (Natural Language Generation) engines, enhancing the chatbot's language processing capabilities.
Pros:
Cons:
Description: Developing a chatbot from scratch using Python involves creating a custom solution tailored to your specific needs. This approach allows for maximum flexibility and customisation but requires significant time, resources, and expertise.
Pros:
Cons:
Description: A pre-packaged chatbot solution is an out-of-the-box chatbot that can be quickly deployed with minimal customisation. This approach offers the most convenience but may not provide the flexibility or advanced features available with other options.
Pros:
Cons:
Here are some of the leading ChatBot vendors:
When selecting a ChatBot vendor, it's important to consider factors such as the vendor's pricing model, features and functionality, customisation options, and integration capabilities. Additionally, it's important to consider the vendor's track record in delivering ChatBot solutions to organisations similar to your own.
The Hudson&Hayes ChatBot Delivery approach provide a seven step process for designing, developing, deploying and maintaining a ChatBot.
Activities:
Deliverables:
Activities:
Deliverables:
Activities:
Deliverables:
Activities:
Deliverables:
Example Deliverables 1: User Journey
Example Deliverables 2: Intent Repository for a Procurement ChatBot
Activities:
Deliverables:
In this phase, the chatbot is deployed to relevant channels and integrated with the relevant systems and APIs.
Deliverables:
In this phase, the chatbot's performance is monitored, and the chatbot is retrained based on feedback to improve its accuracy and effectiveness.
Deliverables:
When implementing conversational AI in procurement, pitfalls can arise at every step of the journey. Here are a few examples:
Lack of Education on ChatBot Capabilities: When organisations introduce a ChatBot, it's crucial to educate users on its capabilities and how it evolves over time. Otherwise, users may become frustrated in the early stages if the ChatBot doesn't answer all their questions. It's important to set expectations and communicate the ChatBot's limitations clearly.
Not Focusing on User Experience: It's easy to get caught up in the technical aspects of delivering conversational AI and forget about the user experience. When users engage with the ChatBot, they should have a seamless experience that feels natural and intuitive. This requires a focus on user experience design and testing.
Going Complex Too Soon: Organisations should ease into the use of ChatBots and show the art of the possible. Going too complex too soon can overwhelm users and lead to adoption challenges. By starting with simple use cases and gradually adding complexity over time, organisations can ensure that users are comfortable and confident using the ChatBot.
Poor data quality: Chatbots rely heavily on data to provide accurate and relevant responses to user inquiries. If the data used to train the chatbot is of poor quality, the chatbot may provide incorrect or irrelevant responses to users. For example, in a procurement context, if the data used to train the chatbot is based on fragmented systems with inconsistent supplier data or duplicate records, the chatbot may struggle to provide accurate supplier information to users. This can result in frustration for users and a lack of trust in the chatbot's ability to provide reliable information. Therefore, it is essential to ensure that the data used to train the chatbot is accurate, consistent, and up-to-date.
Lack of domain expertise: Chatbots require a deep understanding of the domain they are intended to serve. If the team building the chatbot does not have the necessary expertise, they may struggle to create relevant and useful responses. This can lead to a chatbot that is frustrating for users and fails to achieve its objectives.
Insufficient testing: Chatbots require extensive testing to ensure that they are working as intended. This includes both functional testing (i.e., ensuring that the chatbot responds correctly to user inputs) and performance testing (i.e., ensuring that the chatbot can handle the expected volume of traffic). Without sufficient testing, the chatbot may fail to meet user needs and expectations, leading to poor adoption rates and user dissatisfaction.
Maintenance: Like any other technology, ChatBots require ongoing maintenance to perform optimally. They need to be re-trained periodically with new data to ensure that they continue to understand and respond to user messages accurately. Organisations should have a plan in place for maintaining and updating their ChatBot to ensure that it continues to deliver value over time.
Conversational AI is rapidly transforming many industries, and procurement is no exception. Despite the fact that procurement spends a large proportion of time dealing with queries from the business that people could have completed themselves, the use of chatbots and conversational AIs has yet to take off. With the implementation of ChatBots, procurement can benefit from improved user experience, increased productivity, ease of business with suppliers, and increased effectiveness for procurement staff. The use of ChatBots and conversational AIs in procurement is expected to significantly grow over the coming years, providing benefits for procurement, budget holders, and suppliers.
© Hudson & Hayes | Privacy policy
Website by Polar
