Growing the Xcelerate team and keeping the AI conversation open

The UAE added around 250,000 new companies in 2025, taking the national total beyond 1.4 million, and employers continue to expand their teams to serve that growth. Ourselves, at Xcelerate, a digital transformation consultancy and HubSpot Platinum Solutions Partner, has grown within the same market from a founder-led business to a team of seven across the UAE, South Africa, and the Philippines.

This update describes how that growth has unfolded, how we apply AI inside the business, and why we keep an open dialogue with the researchers who study these questions.

A market that keeps adding companies and people

On 7 January 2026, Abdulla bin Touq Al Marri, the UAE Minister of Economy and Tourism, told a media briefing that roughly 760,000 companies have joined the market since September 2021, including about 250,000 in 2025 alone. Hiring has followed that expansion. ManpowerGroup, the global workforce solutions company, reports a UAE Net Employment Outlook (the share of employers expecting to hire minus the share expecting to cut) of +41% for the fourth quarter of 2026, a rise of 24 points on the previous quarter. The same survey finds that 51% of UAE employers have increased headcount compared with a year ago, a share that reaches 74% among very large organisations. Early-career hiring is rising at nearly half of UAE organisations, whilst skills shortages remain the leading cause of delays in filling vacancies, cited by 30% of employers.

How the Xcelerate team has grown

Our own growth follows that picture. Over recent months, we welcomed two consultants and a developer and implementation specialist, joining the colleagues who lead our marketing, content, and delivery work. The team spans three countries and meets periodically in person in Dubai. Our CEO, Ryno Uys, estimates that seven people now deliver what would have required 12 or more a few years ago.

Persistent skills shortages across the market reinforce our approach to hiring. We look for curiosity and grit first, because a curious person with the right way of thinking learns the technical detail quickly, and AI shortens that learning curve further. Certifications in AI carry a short shelf life; prompt engineering dominated two years ago, whilst agent set-up, context, and the MCP layer matter more today. For developers, we look for openness to AI as a partner, with clear rules on how code is produced.

Applying AI inside the business

The UAE sits at the front of the global AI curve. The Microsoft AI Economy Institute, the research arm of the technology company, reports that 73.3% of the UAE's working-age population used generative AI in the second quarter of 2026, the highest share of any economy it tracks. A separate report by AWS, the cloud division of Amazon, and the UAE Artificial Intelligence Office, developed by the UK research firm Strand Partners, puts business adoption at 72%, up from 53% a year earlier, with 31% of those adopters using advanced applications. Wide adoption coexists with shallow depth, and that gap defines where we focus our own effort.

Where AI saves time

Three areas account for most of our use: research, repeatable processes, and project management. In practice, four routines carry most of the benefit:

1. An agent reviews our client channels (messaging, Teams, and email) and delivers one concise digest each morning.
2. Every call is recorded, transcribed, and stored in HubSpot, which leaves our consultants fully present in client meetings.
3. Proposal drafting moves faster, because an agent applies our way of working to the consolidated client feedback.
4. Solutioning gains a sparring partner, since a consultant can check platform capabilities in minutes instead of reading documentation.

Where people decide

Our guiding principle is to outsource tasks and keep the thinking. AI tends to move straight to solutions, so we begin every engagement with process mapping and a clear mind, recording how work happens, who does it, and which systems (named or unnamed) support it. Experience then acts as the final check. One proposed design depended on advisers tagging each meeting type by hand, and our CEO recognised the pattern from a warehouse implementation early in his career: any step that relies on a person remembering to perform it eventually breaks. The colleague tested the idea, saw the result first-hand, and now carries the lesson forward.

How training builds judgement

Each person configures AI around their own way of working, so templates transfer poorly. We train through comparison instead: a new colleague and an experienced leader build the same deliverable in parallel, and we review both versions together, discussing why each section sits where it does and where each insight originates.

Recent research supports the method. KPMG, the professional services network, and the University of Texas at Austin studied 523 early-career professionals completing business tasks with an AI agent, and found that employees with nearly identical knowledge and skill produced dramatically different results once AI entered the workflow. The researchers sorted participants into three groups (AI apprentices, AI delegators, and AI amplifiers), and foundational attributes such as critical thinking, AI literacy, and domain knowledge proved unreliable predictors of which group a person joined.

Two further studies sharpen the picture. Microsoft Research and Carnegie Mellon University surveyed 319 knowledge workers and associated higher confidence in AI with less critical thinking, and higher self-confidence with more. A preprint based on 14 interviews with software engineers in South Korea found that as generative AI absorbed lower-complexity tasks, the productive struggle that once taught juniors disappeared. Ryno shares that concern, and our parallel drafting restores part of that struggle by asking each colleague to build first and compare afterwards.

Guidance for regulated clients

Many of our clients operate in healthcare and financial services, so we raise two considerations early. The first is data residency: Article 13 of UAE Federal Law No. 2 of 2019 generally keeps health data from services provided in the UAE inside the country unless a health authority resolution allows otherwise, and a Ministry of Health and Prevention resolution later set out specific exceptions. The second is the distance between prototype and production: AI is an excellent prototyping tool, and production systems need a security review by experienced people. Organisations should confirm their specific obligations with qualified legal counsel.

A standing dialogue with academics, and what comes next

Technology moves faster than most organisations can reflect on it, so an outside perspective carries great value. This month, we contributed to a doctoral research study on human behaviour and AI usage in organisations, sharing our experience of where AI helps and where it demands caution. The researcher will play findings back to participants, and we plan to reconnect around the end of the year to compare notes. Each exchange tests our assumptions against evidence gathered across many organisations, and the research gains a practitioner's view of daily work.

Over the coming months, we are building a blueprint and manual documenting how Xcelerate works, so that every new colleague inherits the same standards, frameworks, and ways of thinking. The team will keep growing at a pace that protects those standards. We will continue to share what we learn with clients, colleagues, and the academic community.