Executive Summary

Paula Setton Munhoz; Natani Carolina Silveira Cônsolo

Summary prepared by the ResumeAI tool, an artificial intelligence solution developed by Instituto Pecege focused on synthesis and writing.

The current scenario of laboratories seeking accreditation and maintenance of the ISO/IEC 17025:2017 standard is marked by significant challenges, especially due to excessive reliance on manual processes. This reality compromises operational efficiency, increases the risk of errors, and limits the strategic potential of the quality management system. The ISO/IEC 17025 standard, established by the International Organization for Standardization (ISO, 2017), defines international competence requirements for testing and calibration laboratories, being crucial for the validity and reliability of results. However, the documentary complexity and manual consolidation of information create an environment conducive to rework and inconsistencies, as highlighted by Panagiotidou, Chountalas, and Magoutas (2024).

The analysis of the situation in a participating laboratory revealed that approximately eight professionals dedicated, on average, twelve hours per week to administrative and data consolidation activities related to ISO/IEC 17025, representing about 35% of their work time. Of these activities, it was estimated that 65% were repetitive and low-technical-value tasks, such as document organization and spreadsheet updates. Annually, this totaled approximately 3,900 hours dedicated to processes that could be automated, highlighting a substantial loss of productivity and a significant operational bottleneck (CECILIO; DOMINGOS, 2024).

The reliance on manual processes resulted in several operational and managerial failures. Calibration deadlines were missed without automatic alerts, mandatory documents became outdated due to a lack of systematic review, and certificates were archived without adequate critical analysis. Furthermore, teams operated without structured training direction, and the detection of operational failures was inadequate. The laboratory’s official records indicated an average of ten non-conformities per year, but these were generally discovered only during external audits or occasional managerial reviews, suggesting an under-detection of internal problems.

Studies on quality management systems indicate that the actual rate of non-conformities can be three to five times higher than formal records, due to a lack of proactive detection. With an estimated error rate between 10% and 15% in manual consolidations, it was projected that the laboratory was exposed to between 25 and 40 undocumented operational failures annually. This scenario represented a considerable risk to the credibility of the results issued and to the laboratory’s accreditation itself, directly impacting customer confidence and regulatory compliance.

The financial consequences of this operational model were significant. The estimated cost for corrective investigations, potential rejection of results, and the risk of suspension or loss of accreditation ranged from R$ 40,000 to R$ 60,000 annually. The loss of productivity resulting from hours dedicated to manual processes was estimated between R$ 400,000 and R$ 500,000 per year. Additionally, rework generated by inconsistencies in records cost between R$ 15,000 and R$ 25,000 per quarter, adding to the hidden and direct costs of inefficiency.

The main cause of these problems lay in the lack of integrated automation, the excessive use of isolated spreadsheets, and the reliance on individual records, without automatic failure detection mechanisms. This situation generated a higher probability of human errors, inconsistencies in records, and, critically, the non-detection of problems before external audits. As a strategic impact, this scenario compromised operational efficiency, weakened the reliability of the management system, increased regulatory risk, and limited the laboratory’s competitiveness in the market.

The central proposal of this business case was the implementation of an artificial intelligence (AI) training program specifically aimed at laboratory processes. The objective was to optimize compliance with the requirements of the ISO/IEC 17025 standard and, crucially, to improve the proactive detection of non-conformities. The solution aimed to integrate AI and automation tools to optimize routines, interpret data, and consolidate managerial information, transforming the reactive approach to a proactive stance in quality management.

The recommended solution consisted of creating a distance learning (EAD) training focused on the practical application of Generative Artificial Intelligence in laboratory processes. This approach was conceived as a lever for innovation and efficiency, exploring technologies capable of automating content creation, simplifying processes, performing intelligent information analysis, and automatically detecting data anomalies. The training was structured to be applied to real laboratory problems, ensuring relevance and immediate applicability.

The differential of the proposal in relation to traditional training resided in its applied approach oriented towards concrete laboratory problems. Instead of limiting itself to a conceptual exposition on digital transformation, the course focused on the use of AI in existing critical routines, such as critical certificate analysis, document control, indicator consolidation, inconsistency detection, and team training. This practical methodology allowed the incorporation of technology into daily operations, promoting immediate gains in efficiency and standardization.

Generative technology, which includes tools capable of producing texts, presentations, images, videos, codes, and analytical reports, has been fundamental in expanding the capacity of technical teams to structure information and support decisions with greater agility and precision. For the purposes of this business case, three main fronts of AI application stood out. These fronts were designed to act in an integrated manner, covering different aspects of laboratory management and compliance with ISO/IEC 17025, ensuring a holistic and effective approach.

The first application front involved conversational tools for the generation and organization of technical content and reports. Technologies such as ChatGPT were used to assist in structuring critical analysis reports, organizing findings, generating standardized training content, and drafting managerial meeting minutes. The application of these tools allowed for a significant reduction in information consolidation time and ensured the standardization of technical communications between teams, a critical aspect for compliance with the ISO/IEC 17025 standard.

The second front encompassed data analysis and visualization tools, essential for supporting decision-making and anomaly detection. The consolidation of management system indicators, the monitoring of non-conformity trends, and the identification of failure patterns demanded an integrated analysis of multiple databases. Specialized tools were implemented to transform raw data into executive dashboards, allowing management to identify failure patterns, track the evolution of team competencies, and, crucially, detect anomalies and deviations before they became critical problems.

The third front consisted of no-code automation platforms for the integration of administrative and technical processes. Calibration deadline management, document control, skills tracking, and inconsistency detection were recurring processes that frequently required manual intervention. The automation platforms were applied to create integrated workflows that trigger automatic calibration expiration alerts, update training records in real-time, consolidate technical documentation in centralized repositories, and implement automatic validations that identify data inconsistencies in real-time.

This integration eliminated process disruptions, significantly reduced the dependence on manual coordination between teams, and, more importantly, implemented controls that detected failures before they became recordable non-conformities. The combined application of these three fronts enabled the laboratory to evolve from a model based on isolated activities to a model of integrated and intelligent processes, where technology supports both daily operations and strategic quality analysis, and where anomalies are detected proactively rather than reactively during audits.

The implementation of the proposal was structured in three main stages, aligned with the SARIE model (Situation, Analysis, Recommendation, Implementation, Evaluation). The first stage, Situation and Analysis, involved a structured diagnosis of laboratory processes, identifying critical manual activities, mapping data flows, and analyzing failure points. This phase also included stakeholder engagement to identify where failures were not being detected and recorded, providing a solid foundation for the subsequent project stages.

The second stage, Recommendation and Implementation, comprised the structuring and offering of EAD (Distance Learning) training, with practical modules applied to real laboratory problems. This phase included the implementation of AI tools in the three focal processes: critical certificate analysis, team training, and critical document analysis with automatic inconsistency detection. The configuration of monitoring dashboards and automatic alerts was also carried out, ensuring that the technology was integrated functionally and strategically into the laboratory’s daily operations.

The third stage, Evaluation, consisted of the systematic collection of data on Key Performance Indicators (KPIs), qualitative team feedback, monitoring of non-conformity detection, and continuous adjustments to the implementation. Those responsible for the implementation included the quality manager, the IT team, the training coordinator, and an AI specialist. Necessary resources encompassed access to AI platforms, such as ChatGPT Pro, data analysis tools, and no-code platforms, in addition to specialized technical support, team time dedication, and data infrastructure.

The financial feasibility analysis considered both direct costs and projected operational gains for a medium-sized laboratory, with approximately 15 to 20 professionals involved in quality management and ISO/IEC 17025 compliance processes. The initial investment for Artificial Intelligence and Automation tools, considering the integrated use of the three fronts by an initial core of ten professionals, was estimated at US$ 750 to US$ 1,100 monthly, which corresponds to R$ 3,750 to R$ 5,500.

The structuring and offering of the customized distance learning course represented a unique investment of R$ 15,000. Specialized technical support for the first three months was R$ 5,000. Thus, the total estimated investment in the first year ranged from R$ 85,000 to R$ 100,000. In subsequent years, the operational cost would be reduced to approximately R$ 45,000 to R$ 66,000 annually, covering only digital tools, which demonstrates a significant reduction in costs after the initial investment in training and infrastructure.

In the short term, between three and six months, the expected return was associated with a significant reduction in time dedicated to recurring administrative activities and an improvement in fault detection. A reduction of 40% to 50% in time spent on activities such as critical certificate analysis, document updates, preparation of management minutes, and training preparation was projected. For a laboratory with a quality management team of five people, the projected savings were approximately R$ 90,000 to R$ 112,500 annually, representing an annual saving of 600 to 750 hours.

The projections indicated that this value would significantly exceed the initial investment, generating a payback between seven and fourteen months. This rapid recovery of the initial investment is a strong indicator of the proposal’s financial viability, making it attractive to management. The ability to free up team time for higher value-added activities directly contributes to improved productivity and focus on strategic initiatives, rather than repetitive, low-technical-value tasks.

In the medium term, between six and twelve months, a substantial improvement in the proactive detection of non-conformities was expected. With the implementation of automatic controls and anomaly detection, it was estimated that the laboratory would be able to identify between 60% and 80% of potential failures before they become formally recordable non-conformities or before being discovered by external auditors. This ability to anticipate problems is crucial for avoiding correction costs and preserving the laboratory’s reputation.

The total expected economy in the medium term, considering the reduction in time and avoidance of costs for investigating non-conformities (R$ 30,000 to R$ 50,000), totals approximately R$ 120,000 to R$ 162,500 annually, generating a Return on Investment (ROI) of 118% to 149% in the first year. This robust ROI projection reinforces the financial soundness of the proposal, indicating that the adoption of artificial intelligence is not just a measure of efficiency, but a strategic investment with high return potential.

Beyond quantitative gains, the implementation generated important qualitative benefits, such as increased management system reliability, improved process standardization, reduced regulatory risk, and strengthened the laboratory’s competitive position in the market. Process digitalization and automation, as pointed out by Cecilio and Domingos (2024), provide relevant gains in productivity and efficiency, while the absence of integrated automation results in increased errors and operational overload, corroborating the need for the proposed intervention.

Practical tests conducted during the pilot implementation indicated significant potential for efficiency gains. By measuring activities such as critical certificate analysis and indicator consolidation, initially performed manually and then with the support of AI tools, a reduction of 40% to 50% in the time dedicated to these tasks was estimated. Starting from a baseline scenario of approximately forty hours per month per professional, a decrease to about twenty to twenty-four hours per month was projected, freeing up valuable time for higher value-added activities.

Regarding documentary rework, the initial diagnosis identified an average of six monthly occurrences of documents that needed to be redone due to inconsistencies in filling or standardization. During the pilot tests, it was observed that AI tools contributed significantly to automatic standardization and error detection. Based on these observations, a reduction of 60% to 70% in these occurrences was estimated, projecting a drop to approximately two to three monthly cases, which represents a substantial improvement in the quality and efficiency of the documents.

The most critical indicator, proactive detection of non-conformities, also showed improvement. Historically, the laboratory recorded between eight and ten annual non-conformities, generally discovered only during external audits. Tests performed with the automatic detection and pattern analysis functionalities demonstrated the ability to identify anomalies in real-time. Based on these preliminary results, it was estimated that between 60% and 80% of potential operational failures could be proactively detected, representing the early identification of approximately fifteen to twenty-five anomalies throughout the year.

For the preparation time of managerial critical analysis, the tests showed a substantial reduction in data collection and organization effort. Comparing the traditional manual method with the use of AI tools for information synthesis and report generation, a reduction of between 55% and 65% was projected, decreasing from approximately eight hours per month to three to four hours. This optimization allows management to dedicate more time to strategic analysis and informed decision-making, instead of focusing on operational data collection tasks.

During the pilot implementation, the tools were tested in three critical processes of the management system. In the critical analysis of calibration certificates, a time reduction of between 55% and 65% was observed, decreasing from about sixteen hours to six to eight hours per batch. In the development of training materials, a reduction of between 60% and 70% was verified, going from twenty hours per month to six to eight hours. In the critical documentary analysis with anomaly detection, a reduction of between 55% and 65% was estimated, from approximately twelve hours per month to four to six hours.

The observations made during the pilot period confirmed that the implemented tools effectively contributed to the reduction of repetitive activities and the strengthening of control mechanisms. The automatic detection of inconsistencies proved particularly relevant for decreasing the risk of discovering problems only during external audits, which provides greater security and reliability to the management system. This proactive capability for identifying failures is one of the pillars for maintaining accreditation and for the continuous improvement of quality.

The managerial contribution of the proposal lies in the transformation of the laboratory’s operational model, moving from reactive management to proactive, data-driven management. AI training and process automation not only optimize compliance with ISO/IEC 17025 but also free up human resources for higher strategic value activities. This allows the team to focus on innovation, continuous improvement, and the development of new competencies, raising the level of professional engagement and satisfaction.

The structured adoption of artificial intelligence represents a viable strategic investment, capable of generating significant efficiency gains and strengthening the reliability of the laboratory management system. The proposal demonstrated potential to strengthen internal communication processes and team training, allowing for more agile creation of training materials and structured knowledge dissemination. This initiative is fundamental for the continuous development of team competencies and for the laboratory’s adaptation to the demands of an increasingly digital and competitive environment.

In summary, the business case demonstrated that the implementation of an artificial intelligence training program, focused on automating administrative routines and strengthening managerial critical analysis, is a solution with proven operational and financial viability. The projected gains, both in time and cost reduction and in the improvement of proactive detection of non-conformities, justify the approval of the proposal as a priority initiative. This digital transformation not only enhances the laboratory’s efficiency and competitiveness but also ensures the sustainability and credibility of its results in an ever-evolving market.

Bibliographic References:

CECILIO, E. R. de S.; DOMINGOS, J. C. Automação de processos em organizações: um estudo de caso. SemeAd – Seminários em Administração, v. 27, São Paulo: Universidade de São Paulo, 2024. ISSN 2177-3866.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION – ISO. ISO/IEC 17025:2017 – General requirements for the competence of testing and calibration laboratories. Geneva: ISO, 2017.

PANAGIOTIDOU, S.; CHOUNTALAS, P.; MAGOUTAS, A. The multifaceted impact of ISO/IEC 17025 accreditation: a sector-specific analysis in civil engineering testing and calibration laboratories. TQM Journal, v. 37, n. 4, p. 1-19, 2024. DOI: 10.1108/TQM-10-2023-0347.

Executive summary from the Final Course Work of the Specialization in Executive Leadership and Management of the MBA USP/Esalq

To learn more about the course, click here and access the MBX Academy platform