Research questions in software assurance and digital quality
This section organizes technical questions, educational methods, and research directions. Work is identified by its actual status. A practitioner article explains an argument. A proposed study describes a method to be carried out. A completed study requires reported methods, results, limitations, and accessible supporting material where sharing is permitted.
The educational examples on CSVtoCSA are not evidence of measured project outcomes. They help make methods understandable and testable.
Filter the publication and research record
External publication record
Validating the Future series
Five self-published practitioner articles on Sandip’s public LinkedIn profile. These records are not represented as peer reviewed.
Practitioner article · Self-published on LinkedIn · Series part 1
Validating the Future — Part 1: From Simulation to Reality: Why AI Systems Need Continuous Validation
Explains why pre-release evidence is only the starting point for AI assurance and why post-deployment performance, drift, and context changes require a continuous validation loop.
- Author
- Sandip Thorat
- Publisher / date
- LinkedIn · 2026
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- continuous validation, AI assurance, model monitoring, regulated life sciences
- Author contribution
- Conceptualization, practitioner analysis, writing, and publication.
- Suggested citation
- Thorat, S. (2026). “Validating the Future — Part 1: From Simulation to Reality: Why AI Systems Need Continuous Validation.” LinkedIn.
Practitioner article · Self-published on LinkedIn · Series part 2
Validating the Future — Part 2: The Evolution of Validation for Intelligent Systems
Examines how validation planning, testing, change control, and monitoring apply when system behavior can change with data, configuration, and operating context.
- Author
- Sandip Thorat
- Publisher / date
- LinkedIn · 2026
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- intelligent systems, validation lifecycle, changing software, validation approach
- Author contribution
- Conceptualization, practitioner analysis, writing, and publication.
- Suggested citation
- Thorat, S. (2026). “Validating the Future — Part 2: The Evolution of Validation for Intelligent Systems.” LinkedIn.
Practitioner article · Self-published on LinkedIn · Series part 3
Validating the Future — Part 3: How Generative AI Is Beginning to Transform Testing Itself
Discusses how generative AI changes test design, scenario generation, and review of test results while preserving accountable human judgment and challenge.
- Author
- Sandip Thorat
- Publisher / date
- LinkedIn · 2026
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- generative AI, test design, validation records, human oversight
- Author contribution
- Conceptualization, practitioner analysis, writing, and publication.
- Suggested citation
- Thorat, S. (2026). “Validating the Future — Part 3: How Generative AI Is Beginning to Transform Testing Itself.” LinkedIn.
Practitioner article · Self-published on LinkedIn · Series part 4
Validating the Future — Part 4: Continuous Validation & Autonomous Quality Systems
Connects monitoring, feedback, testing, and governance into an operating model for quality systems that increasingly make or recommend decisions autonomously.
- Author
- Sandip Thorat
- Publisher / date
- LinkedIn · 2026
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- continuous validation, autonomous quality systems, monitoring, governance
- Author contribution
- Conceptualization, practitioner analysis, writing, and publication.
- Suggested citation
- Thorat, S. (2026). “Validating the Future — Part 4: Continuous Validation & Autonomous Quality Systems.” LinkedIn.
Practitioner article · Self-published on LinkedIn · Series part 5
Validating the Future — Part 5: Algorithmovigilance: How to Monitor and Control AI Systems in the Real World
Frames algorithmovigilance as a practical post-deployment discipline for detecting drift, anomalous behavior, control degradation, and conditions requiring human intervention.
- Author
- Sandip Thorat
- Publisher / date
- LinkedIn · 2026
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- algorithmovigilance, model drift, post-deployment monitoring, AI controls
- Author contribution
- Conceptualization, practitioner analysis, writing, and publication.
- Suggested citation
- Thorat, S. (2026). “Validating the Future — Part 5: Algorithmovigilance: How to Monitor and Control AI Systems in the Real World.” LinkedIn.
CSVtoCSA
Practitioner articles
Practitioner article · Self-published
The AI Assurance Release Decision: From Model Scores to a Controlled GxP Use
How to combine AI performance, controls, human review and monitoring into a decision about a defined use.
- Author
- Sandip Thorat
- Publisher / date
- CSVtoCSA · 2026-09-05
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- AI Assurance, computerized systems assurance, regulated life sciences
Practitioner article · Self-published
The SaaS Version Problem: Assuring Software You Cannot Freeze
How to assess a supplier release, identify affected functions and choose the evidence needed for continued use.
- Author
- Sandip Thorat
- Publisher / date
- CSVtoCSA · Date not recorded
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- Supplier & SaaS, computerized systems assurance, regulated life sciences
Practitioner article · Self-published
Why “The Vendor Validated It” Is Not an Assurance Conclusion
How to decide which supplier evidence supports your configuration and which questions still need local evidence.
- Author
- Sandip Thorat
- Publisher / date
- CSVtoCSA · Date not recorded
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- Supplier & SaaS, computerized systems assurance, regulated life sciences
Practitioner article · Self-published
When Exploratory Testing Produces Better Evidence Than a Script
When exploratory testing can reveal important failures, and how to keep its findings reviewable.
- Author
- Sandip Thorat
- Publisher / date
- CSVtoCSA · Date not recorded
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- Testing, computerized systems assurance, regulated life sciences
Practitioner article · Self-published
Fewer test scripts do not automatically mean better assurance
How to judge a testing change by risk coverage and evidence quality instead of script counts.
- Author
- CSVtoCSA editorial
- Publisher / date
- CSVtoCSA · 2026-09-12
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- Testing, computerized systems assurance, regulated life sciences
Practitioner article · Self-published
A screenshot is evidence only when it answers a question
How to choose evidence that connects a test claim, the conditions and the observed result.
- Author
- CSVtoCSA editorial
- Publisher / date
- CSVtoCSA · 2026-09-12
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- Testing, computerized systems assurance, regulated life sciences
Practitioner article · Self-published
An AI reviewer can save time and still weaken a review
How to evaluate an AI-assisted review using missed issues, false findings and the final human decision.
- Author
- CSVtoCSA editorial
- Publisher / date
- CSVtoCSA · 2026-09-12
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- AI Assurance, computerized systems assurance, regulated life sciences
Practitioner article · Self-published
Supplier evidence does not remove local accountability
How to reuse supplier evidence while retaining responsibility for local configuration, records and changes.
- Author
- CSVtoCSA editorial
- Publisher / date
- CSVtoCSA · 2026-09-12
- Publication status
- Public practitioner article; not peer reviewed
- Keywords
- Supplier & SaaS, computerized systems assurance, regulated life sciences