Table of Contents
- In-house AI Expertise: The Competitive Edge for SMEs
- The AI Skills Gap in SMEs: An Overview
- Where Does Your IT Team Stand? Competency Analysis as a Starting Point
- Practical Training Strategies for IT Teams
- The Four Essential AI Skills for IT Teams
- The Multipliers Principle: Your AI Center of Excellence
- From Plan to Practice: A Mid-Sized Success Story
- Measurable Success: KPIs for Developing AI Competence
- Common Obstacles and Their Solutions
- Next Steps: Your Roadmap to AI Competence
- Frequently Asked Questions About Building AI Expertise
In-house AI Expertise: The Competitive Edge for SMEs
The current landscape for small and medium-sized enterprises (SMEs) is defined by a widening AI skills gap. According to a 2024 study by Bitkom Research, 78% of SMEs feel inadequately prepared for AI-driven transformation, even though 92% see AI as critical to their future competitiveness.
While large corporations are setting up dedicated AI teams, SMEs face a different reality: How do you upskill your existing IT teams with relevant AI skills?
External consultants can provide short-term support. But let’s be honest—when it comes to sustainable transformation and lasting cost benefits, building in-house expertise is irreplaceable in the long run.
The AI Skills Gap in SMEs: An Overview
The Forrester Research study “AI Adoption in Midsized Businesses” (2024) provides striking figures: Companies with well-trained internal AI capabilities achieve a 34% greater productivity boost and a 29% faster implementation time for AI projects than those who only rely on external service providers.
The costs of lacking AI expertise are substantial. The German Institute for Economic Research (DIW) estimates that the productivity loss from failing to implement AI averages 11.3% each year for SMEs—a painful competitive disadvantage that grows every quarter.
Your existing IT teams have an invaluable advantage: they already know your systems, business processes, and data structures—knowledge no outside expert can match. This blend of domain expertise and newly acquired AI skills is your true trump card.
Where Does Your IT Team Stand? Competency Analysis as a Starting Point
Before investing in further training, you need clarity. What AI skills are already present? Where are the biggest gaps? The good news: Many skills your IT employees possess already provide a solid foundation for AI projects:
- Programming knowledge in Python or JavaScript
- Understanding of data structures and APIs
- Experience with cloud services
- Familiarity with your specific system landscape
A survey by Initiative D21 (2024) also identified common skill gaps in IT teams without an AI focus:
- Lack of conceptual understanding of machine learning and neural networks (78%)
- Little or no experience with prompt engineering (91%)
- Uncertainty regarding ethical and legal aspects of AI use (84%)
- Limited knowledge of how to integrate AI models into existing applications (72%)
For an effective competency analysis, we recommend a three-step approach:
- Self-assessment: Have team members evaluate their own abilities in relevant AI domains
- Skills audit: Conduct hands-on exercises to test different AI competencies
- Gap analysis: Compare current and required skills for your particular AI initiatives
This inventory forms the foundation for a bespoke training plan that is neither too ambitious nor too basic.
Practical Training Strategies for IT Teams
Developing AI skills requires a learning strategy that fits your IT team’s daily routine. Traditional multi-week courses that pull staff entirely away from ongoing operations? Rarely realistic.
According to a 2024 MIT Sloan Management Review study, modular, hands-on learning approaches are up to 3.7 times more effective than purely theoretical programs. The Learning & Development Research Center recommends the following mix:
Modular Learning Paths by Role
- For developers: Focus on practical AI integration, API usage, and prompt engineering
- For data specialists: Deep dive into data preparation for AI and model validation
- For IT management: Emphasis on AI project planning, resource allocation, and risk management
Blended Learning Formats
- 50% hands-on project experience
- 20% peer learning in internal workgroups
- 20% guided online courses from specialized providers
- 10% external workshops and conferences
The “learning by doing” approach—using small, manageable AI projects—has proven particularly effective. These can be implemented alongside daily work and deliver quick, visible wins—a key motivational boost for the entire team.
The Four Essential AI Skills for IT Teams
Not all AI skills are equally important for your IT teams. Focus on those competencies that deliver the most value for mid-sized companies:
1. Prompt Engineering
The “Annual State of AI Report” (2024) identifies effective prompt engineering as the number one practical skill for 76% of successful AI implementations. Your teams should learn:
- Fundamentals of prompt formulation
- Advanced techniques such as chain-of-thought and few-shot learning
- Systematic prompt optimization
- Integrating prompt templates into applications
2. ML/AI Fundamentals
What your teams need is not deep mathematical expertise, but a solid conceptual grasp of AI:
- Differences between various AI model types (LLMs, computer vision, etc.)
- Strengths and limits of current AI models
- Basic AI quality criteria like hallucinations, bias, and robustness
- Evaluating commercial AI offerings
3. AI Integration into Existing Systems
One of the most valuable skills is seamlessly integrating AI into your existing infrastructure:
- API-based AI integration
- Retrieval Augmented Generation (RAG) for company-specific applications
- Orchestrating AI workflows
- Performance tuning and cost control
4. Data Security and Governance for AI Projects
Gartner’s “AI Governance for Midmarket” study (2024) shows that 67% of failed AI projects are due to inadequate data security and governance. Train your teams on:
- Data protection-compliant AI usage
- Implementing access restrictions
- Documenting and versioning AI models
- Monitoring and auditing AI solutions
The Multipliers Principle: Your AI Center of Excellence
Instead of trying to turn every employee into an AI expert—which is neither realistic nor necessary—we recommend a multiplier model. The “Deloitte AI Transformation Report” (2024) demonstrates that businesses with an internal AI Center of Excellence (CoE) see a 42% higher success rate for AI projects.
The Multiplier Model
- Identify your AI champions: Select 2–5 people with a keen willingness to learn and strong communication skills
- Intensive training: Invest in their advanced AI education (e.g., 10–20% of their working hours over 3–6 months)
- Knowledge transfer: Set up regular workshops for champions to share their expertise
- Project support: Deploy champions as advisors on other teams’ AI initiatives
In parallel, establish standards and best practices, including:
- Documentation guidelines for AI projects
- Decision trees for model selection
- Quality criteria for AI applications
- Security and compliance checklists
A strong CoE also needs clear communication structures:
- Weekly office hours for AI queries
- An internal knowledge base for best practices
- Regular brown bag sessions on specific AI topics
- An internal forum for peer support
From Plan to Practice: A Mid-Sized Success Story
A mid-sized automotive supplier with 180 staff and an 8-person IT department carried out a structured AI skills transformation in 2023/2024. Their approach can serve as a template:
Implementation Timeline
Phase 1: Preparation (2 months)
- IT team skill assessment completed
- Three AI champions identified and received basic training
- AI vision and use cases defined in collaboration with business units
Phase 2: Pilot Projects (3 months)
- Two manageable AI projects launched (automated document creation and internal FAQ chatbot)
- Learning by doing with weekly coaching
- All lessons learned and challenges were documented
Phase 3: Scaling (6 months)
- Systematic knowledge transfer to the rest of the IT team
- Internal AI platform established
- Rollout of six additional AI applications
- Ongoing skill upgrades via peer learning
Results Achieved
- 83% of the IT team attained solid AI core competency
- 31% time saved on documentation-heavy processes
- ROI of 287% within 12 months
- 68% reduction in external consulting costs
The “AI Dojos”—weekly two-hour sessions where concrete problems were solved collaboratively using AI methods—proved particularly successful. The teams didn’t just build skills; they also developed a shared understanding of both the possibilities and limits of today’s AI technologies.
Measurable Success: KPIs for Developing AI Competence
To gauge the success of your AI training initiatives, consider both skill development and business outcomes. Without clear metrics, you lack the foundation for making adjustments or further investment decisions.
KPIs for AI Skill Development
- Skill Coverage: Percentage of core AI skills covered within the team
- Project Autonomy: Share of AI projects managed without external support
- Implementation Speed: Time from idea to productive AI application
- Knowledge Dissemination: Number of internal AI workshops and participants
KPIs for Business Success
- Process Optimization: Time saved through AI-supported processes
- Cost Reduction: Lower spend on external service providers
- Quality Improvement: Fewer errors in AI-optimized processes
- Innovation Rate: Number of new business initiatives based on AI
Regular performance measurement enables you to keep refining your training strategy. Gartner recommends reviewing and updating your skill development plan on a quarterly basis, based on real results.
Common Obstacles and Their Solutions
Building AI competence in existing IT teams comes with its share of challenges. According to a 2024 survey by the German Association of IT SMEs, these are the ones that crop up most often:
Time Management and Resource Allocation (76%)
Problem: IT teams are already maxed out with day-to-day tasks; additional training seems impossible.
Solution:
- Explicitly allocate a “learning budget” of 10% of working hours
- Break AI training into short (30-45 min) modules
- Link training with actual work projects
- Provide temporary support for routine tasks
Resistance to Change (62%)
Problem: Staff worry that AI may threaten jobs or render their expertise obsolete.
Solution:
- Position AI as an enhancement, not a replacement
- Celebrate and communicate early wins
- Show clear career paths for acquiring AI skills
- Openly address and discuss concerns
Dealing with Legacy Systems (58%)
Problem: Older systems make integrating AI harder and limit its use cases.
Solution:
- Prioritize API-based integration approaches
- Develop middleware solutions to connect legacy systems
- Choose use cases requiring minimal system integration
- Combine modernization steps with AI projects
Lack of Learning Materials for Specific Use Cases (53%)
Problem: Generic AI courses don’t cover unique company requirements.
Solution:
- Combine basic training with tailored in-house workshops
- Encourage peer learning and exchange with other companies
- Build an internal case study library
- Partner with specialized education providers
Next Steps: Your Roadmap to AI Competence
Developing AI expertise within your existing IT teams isn’t a luxury—it’s a strategic imperative. With the right approach, even SMEs without dedicated AI departments can achieve impressive results.
Key takeaways for your path forward:
- You need a clear starting point: Begin with an honest analysis of your current skills
- Focus on practical skills: Prioritize abilities with direct real-world impact
- Establish multipliers: Set up an internal AI Center of Excellence
- Learn by doing: Combine training with tangible projects
- Make success measurable: Define clear KPIs for skill development and business outcomes
For your company, this means:
- Conduct a structured AI skills assessment within the next 4 weeks
- Identify 2–3 potential AI champions in your IT team
- Define 1–2 manageable AI pilot projects in collaboration with business units
- Develop a six-month plan for systematic skills development
- Establish structures for continuous learning and knowledge sharing
Technology is advancing rapidly—but with a solid foundation of AI expertise, your team is well equipped to keep up with change and turn it into a strategic advantage.
At Brixon AI, we’re happy to support you through the full process—from your initial skills audit to the successful roll-out of your AI strategy. Get in touch to discuss your customized path to AI competence.
Frequently Asked Questions About Building AI Expertise
How long does it take for our IT team to build basic AI skills?
With a structured approach, dedicating 10–15% of working hours to training, your team can gain basic AI skills within 3–4 months. Specialized expertise in certain areas usually takes 6–9 months. The key is to start practical project work right away—purely theoretical courses tend to slow the process down substantially.
Which roles in the IT team should be prioritized for upskilling?
Start with developers who already have programming experience, as well as database and integration experts. These roles can put AI skills into practice most quickly. IT project managers should also be included early on, since they bridge technical options and business needs.
Do we really need in-house AI expertise, or are external service providers enough?
The optimal approach combines both. In-house expertise is crucial for identifying use cases, managing external partners, and driving long-term progress. For specific, one-off challenges, external specialists can be brought in. But without basic internal know-how, you’ll eventually become dependent—and end up paying much more for AI projects.
How much should we budget for AI training?
As a rule of thumb, plan on €3,000–€5,000 per IT employee per year for direct training costs. Add about 10% of working hours for learning and practice. The ROI is typically 150–300% within the first year, if new skills directly translate into productivity gains or cost savings.
Which AI applications are best suited for initial projects?
Start with manageable projects that deliver quick wins, such as:
- Automating document-based processes
- Internal knowledge bases with AI chatbot
- Data analysis and report generation
- Small-scale process optimizations with clear value
It’s important that the first projects bring noticeable benefits to your business units and can be delivered within 2–3 months.
How can we prevent our newly trained AI experts from being poached?
Combine training with attractive development opportunities. Offer roles with AI responsibility, encourage internal innovation, and develop competitive compensation models for AI talent. Above all, foster a learning-driven culture so that knowledge is valued. A system for knowledge sharing ensures expertise is spread throughout the organization—not just held by a few individuals.
Do our IT staff need to learn programming to work with AI?
Basic programming skills are helpful, but not mandatory for every role. With today’s no-code/low-code AI platforms, even less technical team members can implement AI solutions. Deeper integration and customization still require programming know-how. Ideally, your team should blend both skill sets.
How can we make sure our new AI skills actually get used on the job?
Incorporate AI application into regular workflows and goals. Make space for experimentation and ongoing discussions about AI use cases. An internal showcase for successful AI projects can further encourage practical application. Especially effective: set up “AI Dojos”—regular sessions where real challenges are solved together using AI methods.
