Berlin, June 2026 – akirolabs, a leader in AI-augmented strategic procurement, has published four peer-reviewed research papers, with a fifth accepted for publication in November 2026, marking an important milestone in the company’s continued investment in applied artificial intelligence for procurement.

The publications reflects akirolabs’ commitment to advancing enterprise AI through original applied research, rigorous scientific methodology, and product innovation. Across the five papers, the company explores how AI can be made more reliable, more context-aware, and more useful for complex enterprise procurement decisions.

The research series covers key areas that are central to the future of AI-native procurement, including trust in AI-generated outputs, strategic value lever prioritization, domain-specific model training, and enterprise knowledge retrieval. Together, the publications demonstrate akirolabs' ability to translate original AI research into enterprise technologies, combining procurement expertise, applied machine learning research, and advanced AI engineering to shape the next generation of strategic procurement systems.

‍

“While foundation models provide a strong starting point, enterprise procurement requires domain-specific reasoning, structured knowledge, and measurable reliability. Our research focuses on translating advances in AI into practical enterprise systems that deliver consistent and trustworthy decision support.” Dr. Oleksandr Kondratiuk

‍

‍The first paper, titled “Stability Evaluation of Confidence Features Across Model and Data Variants in Large Language Models” published in February 2026 in Research and Science Today, focuses on how organizations can better understand and improve the reliability of AI outputs in procurement workflows. The paper introduces the methodological foundations of akirolabs’ Decision Data Flywheel, showing how procurement decisions can create structured feedback that improves future recommendations and system behavior.

Read more

‍In March 2026, akirolabs published its next research paper in IJMADA, titled “Pairwise Comparison Aggregation with Tool-Augmented LLMs for Procurement Value Lever Prioritization.” The paper examines how AI can support procurement teams in prioritizing strategic value levers across complex business scenarios, while preserving human oversight in complex strategic decision-making.

Read more

‍

‍In April 2026, akirolabs published another paper in IJMADA, titled “Domain-Specific Ontology Construction and LLMs Fine-Tuning for Procurement Knowledge.” The research shows how specialized AI models can be trained with focused procurement knowledge to achieve superior performance on domain-specific tasks, reinforcing the company’s view that smarter, more specialized AI can be more effective than relying solely on larger general-purpose models.

Read more

‍

‍The fourth publication, published in May 2026 in Research and Science Today, introduces “Summary RAG: A Multi-Format Document Retrieval System with Document-Level Summarization”, a new approach for helping companies search and use large collections of internal documents with AI. The research demonstrates how enterprise AI systems can preserve document-level context, improve answer relevance, and handle complex business information more effectively.

Read more

‍‍The fifth publication, titled “Performance Evaluation of Large Language Models and Embedding Architectures for Procurement RAG Systems”, was accepted for publication in Journal of Artificial Intelligence in Engineering Practice. The paper addresses the challenge of evaluating large language models and the performance of embedding architectures in procurement-specific retrieval augmented generation systems.

Read more

‍

Across the five publications, a clear theme emerges: akirolabs is not treating AI as a surface-level feature, but as a core capability that must be researched, tested, and built with the realities of procurement in mind. The company’s research program reflects a long-term commitment to developing AI systems that are suitable for enterprise deployment, grounded in scientific methods, and aligned with the way procurement teams actually work.

Collectively, the research establishes a coherent scientific program focused on advancing trustworthy, domain-specific enterprise AI for procurement. The research program serves as the scientific foundation for multiple AI capabilities integrated into the akirolabs platform.

For enterprise procurement leaders, the research series reinforces an important message: the future of procurement technology will be shaped not only by automation, but by intelligent systems that can support strategy, collaboration, and better decision-making at scale.

These publications highlight the maturity of akirolabs’ internal AI research program and the CTO’s role in defining its scientific and technological direction. By integrating original research into enterprise software development, the research and engineering teams continue to advance the company's AI platform while reinforcing its position as a technology leader in strategic procurement.

About akirolabs

Akirolabs is an AI-augmented strategic procurement platform that helps enterprise procurement teams improve category strategy, collaboration, and decision-making. The company brings domain-specific AI, agentic workflows, and structured procurement intelligence into strategic procurement processes, helping organizations modernize how they create strategies, align stakeholders, and make better business decisions. The company combines proprietary AI technologies with ongoing applied research to deliver enterprise-grade decision intelligence for strategic procurement. The company maintains an active applied AI research program that continuously informs the development of its enterprise platform.

‍

‍