Nana Wu
Dwi Wanna
Suro Senen
Evan Lee Salim
Sagoh
In a move that highlights the shifting landscape of enterprise software, Newforma—a long-standing leader in project and information management for the architecture, engineering, and construction (AEC) sectors—has announced the acquisition of Toronto-based startup Lumi AI.
The deal, the financial terms of which remain undisclosed, marks a significant consolidation of agentic analytics technology into a legacy industry vertical. By integrating Lumi’s sophisticated data-processing engine, Newforma aims to transform the massive, often siloed repositories of project data that AEC firms have accumulated over the past two decades into a source of actionable, verifiable intelligence.
Main Facts: The Intersection of Legacy Data and Generative AI
Lumi AI, founded in 2023 by Ibrahim Ashqar and Tudor Boiangiu, carved out a niche as an "intelligent assistant" for enterprise data. While many startups in the AI space focused on creative content generation, Lumi focused on the high-stakes world of business intelligence. Its platform was designed to bridge the gap between complex, messy enterprise data sets and the non-technical users who need to derive insights from them.
Under the terms of the agreement, the transition is both structural and strategic. Tudor Boiangiu will join Newforma as the Vice President of AI Engineering, leading a dedicated team of Lumi engineers who will continue their operations in Toronto. Conversely, co-founder Ibrahim Ashqar is stepping away from the business following the transaction.
For Newforma, the acquisition is a tactical maneuver to accelerate its roadmap for "Vojo," a new AI-powered assistant, and to enhance the analytical capabilities of its existing Konekt cloud platform.
A Chronology of Growth and Exit
The journey of Lumi AI reflects the rapid, often volatile lifecycle of early-stage AI ventures in the post-ChatGPT era.
- 2023: Ibrahim Ashqar and Tudor Boiangiu launch Lumi AI in Toronto with the mission of democratizing data analytics. The platform initially functions as an assistant that helps users interpret data and generate charts, effectively automating tasks that previously required dedicated data science teams.
- Late 2023 – Early 2024: The company evolves into an agentic analytics platform, moving beyond mere interpretation to proactively identifying patterns and actionable insights. The company begins serving a diverse client base ranging from CPG companies to major supermarket chains.
- May 2025: Facing a challenging fundraising environment for mid-tier AI startups and a realization that keeping pace with frontier model development requires massive capital, the founders make the decision to pursue an exit.
- Post-May 2025: The acquisition by Newforma is finalized, securing the future of the technology and the employment of the engineering team while transitioning the product into a specialized AEC-focused tool.
Supporting Data: Why "Build vs. Buy" Won the Day
The acquisition serves as a poignant case study in the "build versus buy" dilemma currently facing established enterprise software firms.
According to Ashqar, the core value proposition of Lumi was never just the AI model itself—it was the "scaffolding." He points to a critical reality in the industry: while large language models (LLMs) like Claude or GPT-4 are powerful, their out-of-the-box accuracy when connected to proprietary enterprise data is often insufficient for business-critical decisions.
"We would show potential clients studies on how to enable self-service analytics," Ashqar explained. "When you connect a raw model to data, you might see 21% accuracy. We built the scaffolding—the infrastructure, the guardrails, and the documentation—that pushed that to over 95%."
The maintenance of this "scaffolding" is the hidden cost of AI adoption. Lumi’s internal testing showed that without constant updates and maintenance, the reliability of such systems can plummet from 95% to 65% in just one month. For a company like Newforma, which manages decades of mission-critical construction documentation, the cost of building that expertise internally would have been prohibitive, both in terms of capital and time.
Official Responses: Prioritizing Verifiability
In the AEC industry, where project delays or errors can lead to massive financial losses and safety risks, the bar for AI implementation is exceptionally high. Newforma’s leadership has been vocal about the necessity of "explainable AI."
"Our customers don’t need another AI that sounds confident; they need answers they can verify," said Newforma CEO Peter Cannone.
The integration of Lumi’s engine is intended to meet three specific criteria:
- Explainability: The AI must show its work, detailing the logical steps taken to reach a conclusion.
- Verifiability: Every insight must be traceable directly back to the source data within the Newforma platform.
- Accuracy: The system must be robust enough to support actual business decisions rather than just acting as a "chat interface."
By embedding these capabilities into Newforma Konekt, the platform will be able to perform high-level diagnostics, such as identifying which consultants are consistently slowing down RFI (Request for Information) turnarounds or flagging when a current project’s risk patterns mirror those of a previous, unsuccessful venture.
Implications: The Reality of the "AI Winter" for Startups
The Lumi AI acquisition provides a sobering look at the current market climate. Despite what many would consider "solid" growth—including triple-digit annual recurring revenue growth and a 109% net revenue retention rate—Lumi found it difficult to raise a $10–$20 million Series A round.
Ashqar was candid about the disconnect between the startup’s performance and market expectations. "We were doing well. But they aren’t the ‘crazy’ growth metrics that are expected from AI companies today," he noted.
This reality creates a "widening gap" for many AI startups. To stay relevant, companies must invest heavily in R&D just to keep up with the pace of innovation from incumbents like OpenAI or Google. When a startup cannot raise the capital required to bridge that gap, an exit to a larger, more established partner becomes the most pragmatic path forward.
A New Chapter for Industry-Specific AI
For the AEC industry, the implications are profound. As the sector continues to digitize, the amount of data generated by project management software has become overwhelming. By applying agentic AI that understands the specific nuances of construction documentation, Newforma is essentially unlocking a new layer of productivity.
The acquisition ensures that Lumi’s technology will not be discarded or lost to the "valley of death" that claims many promising startups. Instead, it will be repurposed to serve a large, established user base that is desperate for the very efficiencies Lumi was designed to provide.
For the employees at Lumi, the deal provides stability. For the stakeholders, it provides a clean exit that avoids the need for further dilution. And for Newforma, it provides a shortcut to a future where project data is no longer a static archive, but a living, breathing asset that proactively alerts project managers to risks before they manifest into costly delays.
As the industry watches, the success of this integration will likely serve as a blueprint for other vertical-specific software companies looking to leverage AI without the grueling process of reinventing the "scaffolding" from scratch. The era of the "generalist" AI startup may be cooling, but the era of the "specialist" AI integration is clearly just beginning.
Pevita Pearce
