Artificial intelligence in professional services is approaching a turning point. Accounting and advisory firms have spent the first phase of adoption focused largely on efficiency: automating routine tasks, reducing manual effort and freeing up time. Abu Bakkar, chief innovation officer at HLB, argues that the next phase must be different. The technology should not only make processes faster; it should make audit and advisory work better.

Bakkar spoke at HLB’s Technology & Innovation Symposium in Barcelona, where member firms gathered for a hands-on programme designed to move AI from abstract discussion into practical application. The event included an AI hackathon using the no-code platform Lovable and document intelligence technology from DataSnipper. Teams were challenged to turn real-world professional services problems into working prototypes.

Key facts

• Abu Bakkar is HLB’s chief innovation officer and leads the network’s technology and innovation strategy.
• HLB held its second Technology & Innovation Symposium in Barcelona after a first edition in London.
• The Barcelona event shifted emphasis from talking about AI to building practical applications.
• The programme included a keynote from Jordi Mompart, director of Barça Vision at FC Barcelona, on AI and decision-making.
• Member firms took part in an AI hackathon using Lovable’s no-code platform and DataSnipper’s document intelligence tools.
• HLB Bové Montero, HLB Spain, shared its AI roadmap, including an audit tool, an automated tax workflow and a firm-wide training programme.
• HLB has a research project with the University of Manchester examining AI adoption across member firms.
• HLB also has a commercial relationship with Lovable to give member firms a route from prototype to production.

The symposium was built around a deliberate shift in emphasis. Rather than spending the day discussing AI in the abstract, HLB wanted delegates to build. The opening keynote from Jordi Mompart, director of Barça Vision at FC Barcelona, illustrated how AI is reshaping decision-making and performance at a major football club without compromising its footballing identity. Bakkar said the session was a useful reminder that AI is not only an accounting firm conversation. It is happening across sectors and across functions.

From there, Lovable set up the afternoon hackathon, and delegates moved directly into application building. The structure was intentional. Teams had time to understand the challenge and the tool, then design a workflow before they began building. The thinking came first; the prototype followed. Bakkar said this helped domain experts approach AI as a practical problem-solving exercise rather than a technical curiosity.

From efficiency to quality

For many professional services firms, AI adoption still centres on efficiency. Firms are streamlining operations, cutting costs and saving people time on routine work. Bakkar does not dismiss that. Efficiency matters, and it remains the fastest and most easily proven return on investment. But he argues that the opportunity is starting to move toward quality. AI can make processes smarter, not just faster, and it can open up genuinely new ways of working.

The symposium was designed to give firms space to test that proposition for themselves. Teams applied AI tools to challenges they face in their day-to-day work, with the aim of identifying ideas that have real potential to be taken forward and implemented across the network. Bakkar said the most important outcome was confidence. Firms should leave feeling that they can innovate even though they are not technologists by trade. They are domain experts, and that is exactly the right starting point.

That message matters because professional services firms often treat AI as a specialist function. In practice, the people closest to the work are usually best placed to identify where friction exists. They know which parts of an audit process are repetitive, which advisory workflows are difficult to scale and where clients need faster, clearer answers. When those experts are given accessible tools, the gap between identifying a problem and building a solution narrows.

Strategic objectives across the network

HLB’s core objective is to use cutting-edge technology to support the delivery of high-quality service. The network also wants to shift the conversation from internal operations toward external growth opportunities. Alongside that, HLB is cultivating a culture of technology leadership and thought leadership across the professional services sector, not just within individual firms but as a network-wide identity.

A single day in Barcelona is one visible expression of that ambition, but it sits inside a broader push. Content development, speaker curation and facilitation are led centrally by HLB International, while member firms host and contribute in turn. The format is designed to build a genuine learning culture, enabling participants to get hands-on, share ideas with peers and build confidence with new technology rather than simply hearing about it. The symposium also brings together the future leaders of member firms, which Bakkar described as arguably as important as anything built on the day itself.

That focus on future leaders reflects a wider challenge in professional services. AI adoption is not only a technology question; it is a leadership question. Firms need people who can spot opportunities, assess risk, bring colleagues along and turn experiments into repeatable processes. A network-wide event gives those emerging leaders a common language and a shared set of reference points.

What teams built in the hackathon

The afternoon hackathon had firms building tools aligned to real-world challenges using Lovable’s no-code platform and DataSnipper’s document intelligence tools. Bakkar said what stood out most was the change in how people approached the problem. There was real creativity from domain experts when they were handed the right tools. These were people who knew exactly where the friction was in their daily work but would not normally call themselves builders.

The hackathon was structured to support that. Teams had time to get familiar with the challenge and the tool, then design a workflow before they started building. Lovable allowed rapid prototyping, so teams could walk away with something that actually worked. The presence of a concrete reference point helped. Earlier in the afternoon, HLB Bové Montero, HLB Spain, walked through its own AI roadmap. It discussed an audit tool, an automated tax workflow and a firm-wide training programme, plus the discipline behind how it tests new platforms.

Its rule is simple: internal capability comes first, and client-facing work only follows once the firm has proven the idea on itself. That mindset carried into the hackathon. Pairing domain experts with accessible technology tools proved to be a reliable way to tackle almost any challenge put in front of a team. The prototypes were not intended to be finished products, but they demonstrated what is possible when the barrier to entry is lowered.

How representative the event is

Bakkar said the Barcelona event is genuinely representative of a shift seen across the network over the past few years. Firms are more open to change and innovation than they have ever been. They are experimenting with AI more than ever, and network leaders are actively championing that change. The symposium is not an isolated exercise in one city for one afternoon; it sits on top of a much wider pattern.

HLB ran a research project with the University of Manchester looking specifically at how AI is being used across member firms. Bakkar said the project gave the network a valuable picture of where adoption is happening and at what scale. Some leading firms are already going beyond automating processes and are building agents that act almost like copilots inside audit and advisory work. Because it is becoming easier to build solutions in-house, more firms are building first and buying off-the-shelf second. More AI-specific job titles are also appearing across the network, with roles designed to sit with domain experts and help embed AI properly into how a firm works.

That trend line matters for firms that are still early in their AI journey. It suggests that adoption is not a single leap but a series of practical steps. Some firms start with document review or data extraction. Others begin with internal training or workflow automation. The common factor is a willingness to test, learn and refine rather than wait for a perfect solution.

From prototype to production

A 90-minute hackathon build is one thing. Turning a prototype into something a firm actually uses with clients is another. Bakkar said the key is lowering the barrier to entry for people who are not technologists by background. A domain expert does not need to become a developer to build something useful anymore. But that does not mean the prototype can remain a workshop novelty.

HLB has a commercial relationship with Lovable that gives member firms a practical route to take prototypes into production. HLB Bové Montero’s roadmap offers a useful template for the discipline that path requires: prove it internally, on real work, before putting it in front of a client. That approach helps firms manage risk, build confidence and understand where a tool adds value. It also prevents experimentation from becoming an end in itself. The goal is not to build the most demos; it is to build tools that colleagues trust and clients can rely on.

That discipline is especially important in audit and advisory work, where quality, documentation and accountability are central. A prototype may show that a workflow can be automated, but production use requires attention to data handling, review processes and regulatory expectations. Firms need to know which tasks can be safely supported by AI and which still require significant professional judgement.

What success looks like

Looking ahead, Bakkar identified several markers of success. First, HLB wants to keep the culture of experimentation alive and make sure ideas from events like the Barcelona symposium turn into live applications running in the business. Second, AI must be used responsibly across the network as adoption scales. Third, the focus should remain on AI augmenting people, not replacing them.

Practically, Bakkar would like the symposium format to keep evolving in the way it already has between London and Barcelona. Different host firms and different local flavour should sit alongside the same underlying ambition: bring the network together, build something real and take it home. If next year’s symposium shows the same kind of jump seen from last year to this one, that would be success in itself.

Challenges and opportunities

Bakkar sees challenges clustering around three areas. On the people side, firms face skillset gaps, a lack of formal training inside firms and real fear around change and job displacement that needs to be addressed honestly rather than dismissed. On the technology side, integration with legacy systems is not always possible. The pace of change in underlying models makes it hard to keep up, and choosing the right tool from an increasingly crowded market is becoming more complex.

Governance is the third challenge. Firms must make sure AI use lines up with internal compliance and regulatory expectations around data. They need to be clear-eyed about which processes are ready to be automated, which use cases make sense and how AI sits alongside the realities of hourly billing. These are not abstract concerns. They affect how firms price work, how they supervise staff and how they demonstrate quality to clients.

The opportunities, however, more than match those challenges. The most widespread and immediately useful applications are still in the efficiency layer: document review, data extraction and drafting first versions of routine correspondence or reports. That is the entry point for most firms, and it is where the return is fastest and easiest to prove. Beyond that, there is real opportunity in AI as an analytical layer on top of audit and advisory work. It can help teams spot patterns or anomalies across much larger datasets than a person could realistically review, sharpening the quality of advice rather than just the speed of producing it.

There is also a talent opportunity. Firms that can show they are genuinely building with AI, not just talking about it, have an edge in attracting the next generation of professionals who expect to work this way. For Bakkar, that connects back to the central argument: AI in professional services should move beyond efficiency and become a route to better judgement, better quality and a more confident culture of innovation.

Source: UKTN News