theBIMtech was established in April 2024 to assist the Architecture, Engineering, and Construction industry (AEC) industry in building efficient Building Information Modeling (BIM) workflows through long-term consulting and coaching.
The company began by offering standardized course services combined with real project support and, over time, developed a comprehensive BIM transformation blueprint that helps its clients select and plan the most suitable implementation paths. Their core client base is small and medium-sized enterprises, and they have witnessed steady growth as more companies seek tailored digital transformation support.
Yet as they helped their clients, they realised that they themselves needed to up their game in order to process efficiently and coherently the large amounts of data they needed to handle. They achieved these capabilities through a partnership with the AI-native workspace illumi.
SMEhorizon speaks with Robert Liu, Co-Founder and CSO of theBIMtech on his company’s foray into AI tools, their partnership with illumi, and the outcomes achieved. Chang LingYi and Andrey Leskov, Founders, illumi, also weigh in on how SMEs like theBIMtech can avoid some of the common pitfalls when starting to work with AI tools.
Managing information with AI tools
Liu recalls that researching topics for theBIMtech’s clients often meant sifting through large volumes of information and spending considerable time before being able to organize it coherently.
“We needed a solution that could streamline this knowledge management process—both for internal team discussions and when developing strategy frameworks for clients,” he explained.
Liu learned about illumi through a friend’s recommendation, as he searched for a tool that could visually map research threads, conveniently compile notes and documents, and provide an AI assistant capable of answering questions and integrating knowledge within the same environment.
“When evaluating illumi against other options, I looked for intuitive usability, versatile note-taking and whiteboarding features, and the effectiveness of their AI assistant in supporting research and business planning,” he explained.
Chang and Leskov explain that identifying where AI can make an immediate impact like theBIMtech has done, is key to escaping what they call the 1% ROI trap. “Most AI pilots fail because companies focus too much on the technology itself rather than how it integrates with people and workflows,” they explain.
“According to BCG, while 92% of companies invest in AI, only 1% achieve mature results. The key issue is that AI success is 70% dependent on culture, processes, and people—not just algorithms.” Some other areas that AI is suited to are repetitive, high-friction tasks like customer service responses or data entry
AI as a team member, not a replacement
The gains for the company have been clear. Liu explains that “with illumi, our team can efficiently brainstorm and explore ideas directly on a digital whiteboard, then swiftly organize key points and conclusions.
“This has dramatically streamlined our business model discussions—where previously, we would need several follow-up meetings to research and organize ideas. Now, we’re saving at least 50% on meeting iteration cycles, allowing us to move from concept to decision much faster.
Yet not all tasks can be delegated completely to AI. Liu shares that for certain specialized use cases, he still needs to use dedicated models or tools as the models available through illumi are, currently, not always as robust.
Yet while Liu feels that further enhancements will reduce his dependence on external, specialised tools, Chang and Leskov caution against over-reliance on AI, as it may lead to “cognitive debt”, the gradual erosion of human problem-solving skills when we over-rely on AI.
“Like financial debt, it accumulates silently,” they explain. “Workers stop thinking critically, lose creativity to challenge assumptions, and depend on AI for even basic reasoning.”
The consequences for companies can be devastating. Chang and Leskov warn of declining innovation as excessive AI use weakens original thinking leading to teams producing generic, shallow outputs instead of breakthrough ideas, as well as false confidence in AI outputs where employees miss out onerrors or biases hidden in polished but flawed responses.
Another danger is weaker collaboration skills when employees default to AI instead of discussing ideas with colleagues. In the last case, they explain, “companies lose the “creative friction” that drives innovation.”
Avoiding cognitive debt
How can businesses avoid these pitfalls while leveraging the potential of AI for their business? From his experience with illumi, Liu suggests treating new AI tools as if they were new team members. “Give clear, specific instructions and monitor the progress closely,” he advises.
“Stay informed about developments in the AI landscape, and understand that the right tool can significantly enhance productivity and knowledge management—but it’s important to match the tool’s strengths to your business’s actual needs.”
Chang and Leskov outline the mindsets companies need in order to collaborate better with AI:
- “AI is a team sport.” Treat AI as a collaborator, not a replacement. The best results come when humans and AI refine each other’s work (e.g., editing AI drafts together).
- Embrace imperfection. Rough, early AI outputs invite team input—polished ones discourage critique. Encourage a culture where questioning AI is normal.
- Prioritize context over prompts. AI performs best when it understands team knowledge, goals, and past decisions (not just one-off queries). Tools like illumi help teams build shared AI memory.
- Reward experimentation. Celebrate failures as learning steps—teams that iterate together adapt faster.
Meanwhile, for SMEs looking to introduce AI into their organisation, they advise:
- Avoiding over-customization. Start with affordable, off-the-shelf AI tools (e.g., ChatGPT for drafting, Canva’s AI for design) before investing in bespoke solutions.
- Focusing on “low-hanging fruit.” Automate repetitive tasks (e.g., email sorting, report generation) where AI gives quick wins.
- Training teams to use AI critically. Teach employees to fact-check AI outputs and refine them—don’t assume AI is always right. Compare and share practices between team members to learn.
- Leveraging collaborative AI tools. We built illumi to help SMEs pool knowledge so AI learns from the team’s context, reducing wasted time on re-explaining tasks.
- Partnering wisely. If lacking in-house expertise, work with AI vendors that offer clear onboarding and support—avoid complex systems requiring heavy IT resources.