49 episodes listed below
Putting data to work inside performance and internal audit.
Conor McGarrity and Yusuf Mulla talk about how auditors use data. Sometimes it is just the two of them, and sometimes they have guests such as algorithm auditor Cathy O'Neil and KNIME data scientist Maarit Widmann. The show has now ended, and its archive covers practical topics like profiling data, testing hypotheses and building tools that auditors can maintain themselves.
PRESS PLAY
Auditing with Data is a podcast about how data is used in performance audit and internal audit. It was previously known as the Assurance Show. Conor McGarrity and Yusuf Mulla host. The conversations are plain and practical, and they focus on what auditors do with data once they have it. Sometimes the two hosts talk on their own. In '54. Analysis' they try to settle what the word 'analysis' means in data analysis, because people often understand that stage of audit data work differently. They cover the patterns they expect to see in transactions, early profiling, rules that test a hypothesis, the difference between matching and blending, and checking exceptions until only true anomalies remain. Other episodes have guests from quite different backgrounds. Cathy O'Neil is introduced as a mathematician turned quant turned algorithm auditor, and as the author of Weapons of Math Destruction. Her point is that most algorithm auditing is human work: working out what could go wrong, how bad it would be and who it would affect. The technical part comes last. One example is inferring gender or race when an insurance algorithm never collected either. That episode was also released as a joint special with the Performance Audit Report. KNIME comes up more than once. Maarit Widmann, a data scientist at KNIME, talks about time series analysis and how much of it is data access, pre-processing and deciding what to do about a missing week of data. She also says to explore the data first, then check whether a model adds anything beyond what exploration already shows. Michael Drechnowicz, head of internal audit at TK Elevator, who previously worked in internal audit at DHL, talks about KNIME as a user. He values having the data, the reconciliations and the findings in one place, so completeness, accuracy and traceability are clear to an external quality audit. He also says KNIME's community portal often had answers to his problems before he ran into them. Erica Toelle, Senior Product Marketing Manager for Records Management and InfoGov at Microsoft, talks about compliance in Microsoft 365 and her book The Compliance Guide for Microsoft 365. Her background is in consulting on records management, including work with the team that designed SharePoint's records features. An episode titled '58. The final episode' explains the decision to stop: Conor McGarrity moved on to a new role, and Risk Insights, which produces the show, shifted its focus toward independent audits and away from internal audit. The same episode includes a passage about a COVID tracker. A central team built it together with an auditor whose R skills were rusty, and that auditor could then maintain the tool inside the audit team. The goal described there is tools any auditor can pick up with minimal training. The show is now a finished archive rather than an ongoing feed.
Made for: Suits an internal or performance auditor who wants to use more data in audit work. It also suits a listener curious how algorithm auditing, KNIME and Microsoft 365 records management look from an audit perspective.
What sets it apart: It keeps data work tied to audit concerns: traceability, completeness and accuracy, and telling ordinary exceptions apart from true anomalies. Its guests include people from KNIME and Microsoft, as well as a head of internal audit and an algorithm auditor.
The podcast for performance auditors and internal auditors that use (or want to use) data. Produced by Risk Insights.
A flagged transaction isn't a finding; Conor and Yusuf make you check assumptions before writing it up.
Press play if you want
Short reviews from the PlayNext desk, based on the episodes we processed.
A practical farewell about auditor training, data self-sufficiency, and the algorithm work ahead.
The last Auditing with Data episode works as both a sign-off and a handoff. Connor's move into the BDO partnership in Australia changes the hosting shape, while Risk Insights' turn toward independent audits explains why the show is giving way to Algorithm Integrity Matters, a…
A grounded look at how TK Elevator's internal audit team uses KNIME for traceable, repeatable analytics.
Internal audit is treated as a working system here, not a narrow compliance function. The episode follows Michael from his finance and accounting background into the broader remit of his team at TK Elevator, where audit work touches business processes from sales to purchasing,…
Algorithm auditing begins with naming possible harm and who bears it, before the statistical test arrives.
Algorithm auditing is framed here as an exercise in judgment before it becomes a matter of measurement. The episode follows an auditor’s shift from finance during the credit crisis into the practical work of asking what might go wrong in an automated system, how serious the harm…
A grounded look at low-code workflows, missing data, and the judgment behind time series analysis.
Data science is framed here through the practical choices that decide whether an analysis deserves trust. Marit Vidman, a data scientist at NIME, explains how visual programming can support data science work without coding, then moves into time series analysis: missing weeks,…
Specialists walking through software's tradeoffs, one system at a time.
Founders and engineers on AI coding agents like Claude Code, and supply chain security.
Auditing with Data closes with Yusuf Mullah saying Connor McGarrity has moved to a partnership role at BDO in Australia, while Risk Insights shifts toward independent audits and a new podcast, Algorithm Integrity Matters.
Michael Drechnowicz, head of internal audit at TK Elevator, describes a 16-auditor team spread across South America, Asia Pacific, Europe, and China, covering about 100 entities.
["Cathy O'Neil says algorithm auditing starts with a plain-English question: does the system work as expected, and who does it fail?", "She traces her skepticism to the 2007 credit crisis, then to data science jobs where her models sorted people by money first, then gender and race.", "The episode connects FICO scores, welfare fraud systems in the Netherlands and Australia, and the UK A-level grading algorithm to the same problem of automated harm.", "She argues that regulators, not just companies, need standards for credit, insurance, housing, and hiring so algorithms can be measured without making the metrics gameable."]
["The Assurance Show episode 55 focuses on KNIME's visual programming model and why Marit Vidman says it helps data scientists, business analysts, and auditors share one workflow.", "Hosts Conor McGarrity and Yusuf Mullah speak with Marit Vidman, a KNIME data scientist and education team member who writes courses, gives presentations, and teaches users.", "Marit discusses a book she is writing on codeless time-series analysis, including Arima, seasonality, stationarity, and missing-data decisions.", "She says new users can start at knime.com through events, learning pages, online courses, webinars, books, and the open-source download."]
The Assurance Show's 2022 discussion examines why "analysis" remains ambiguous in data work for performance and internal auditors.
Yusuf Moola and Connor McGardy wrote The Data Confident Internal Auditor for non-technical auditors facing an “avalanche” of data expectations.
An interview with Erika Telly, Senior Product Marketing Manager at Microsoft, discussing her book on compliance for Microsoft 365 and her background in records management.
An interview with Michael DeCero, internal audit analytics manager at TDS, discussing his career path and the role of data analytics in internal auditing.
An interview with Steve Rommel, a senior analytics manager, discussing his career journey and the use of data analytics in internal audit.
An interview with Ari Levian, CIO and CRO of Peregrine Investment Managers, discussing his dual role and the changing risk landscape.
An introduction to Audit Scotland and its role in public sector auditing, with a focus on performance audits and data use.
An interview with data visualization expert Ali Torban about using data visualization in audit fieldwork, risk assessments, and audience-ready reports.
An interview with Andrew Clark, CTO of Monitor, discussing machine learning integrity and data analytics for auditors.
This episode discusses the principle of maximizing benefits from data use in audit, addressing challenges like access, low-value analysis, superficiality, timing, and false positives.