search
How software, data and AI systems connect: from a bank feed in your accounting software to AI features that send your firm's data to an AI provider.
A client clicks Connect bank account in their accounting software, signs in securely at their bank, and the next morning the latest transactions appear in the books. No one downloaded a statement. No one emailed a spreadsheet. Somehow, two separate systems exchanged exactly the information they needed.
This course explains how that works, starting from first principles and building toward the AI tools now being added to professional software.
Dr. Jason Dean begins with a simple problem: a person can visit a website and retrieve information, but how does one piece of software ask another system for data? That question leads to APIs.
Using plain language and real examples, the course introduces the basic pieces of an API connection: the caller, the server, the API call and the credential. Learners examine an actual request to the Bank of Canada’s public exchange-rate API and the reply that comes back. They see how different request methods are used to retrieve or change information, make an API call from Excel, work through API documentation, and interpret common status codes such as 401, 403 and 429.
The same basic architecture also sits behind many new AI features.
A fictional accounting partner, Maya, is offered a tool that automatically suggests categories for transactions arriving through her clients’ bank feeds. Following that example reveals a subtle change in the direction of the data. The bank connection brings information into the accounting system. The AI feature may send information from that system to an outside AI provider.
Learners examine the request being sent, what information it contains, and what happens after it reaches the model. The course explains how an AI API differs from a conventional data lookup, including generated responses, variation between outputs and token-based pricing. It then extends the same framework to tool use, where an AI model can request that other API calls be made on its behalf.
That provides a foundation for understanding agents, connectors and the Model Context Protocol without treating them as mysterious new technologies. The course also examines prompt injection and how an apparently ordinary piece of external content can influence what an AI system does when that system has access to tools or data.
The final section turns the technical ideas into a practical way of evaluating software connections. For any connection, ask three questions:
What can it reach?
Where does the information go?
What can it do?
Learners apply those questions to permission scopes, consent controls, AI providers’ data terms, subprocessors, data processing agreements, audit logs and reliance on outside services. They then return to Maya’s firm and decide what should be examined before enabling the proposed AI feature across every client account.
Throughout the course, short active-learning exercises ask learners to interpret what they are seeing rather than simply remember terminology. These include distinguishing an API from an API call, reading a rate-limit response, diagnosing stopped bank feeds from their status codes, examining the contents of an AI request, assessing the consequences of a prompt-injection attack, and evaluating the permissions given to connected software.
By the end of the course, learners should be able to:
explain what an API is and distinguish an API from an individual API call;
read a real API request and response and determine whether the request succeeded;
explain how software can receive permission to access bank data without receiving the client’s banking password;
explain how an AI feature sends information through an API and how an AI model can request calls to other tools or services; and
assess a proposed connection between a firm’s data and an outside system by determining what the system can reach, where the information goes and what actions it can perform.
Note: This course provides general educational information. Laws, professional requirements and organizational policies vary by jurisdiction and can change over time. Consult current requirements applicable to your organization and jurisdiction.
Associate Professor of Economics | King's University College at Western
#Dr. Jason Dean **Associate Professor of Economics and Applied Economist** Dr. Jason Dean is an Associate Professor of Economics at King’s University College at Western University and holds a PhD in Economics from McGill University. He teaches university students and working professionals, drawing on experience in academic research, economic consulting, public policy and the practical use of artificial intelligence. His CPD work covers applied economics and emerging AI tools. His approach is practical and evidence-based: explain how something works, identify the assumptions and limitations, and give learners a framework they can use in their own work. Jason has worked extensively with generative AI since the emergence of ChatGPT, incorporating it into teaching, research and professional education. He has delivered three presentations for Haver Analytics, most recently a June 2026 session on Model Context Protocol (MCP) servers. He has also presented on AI for the Canadian Chamber of Commerce and the Toronto Association for Business and Economics. His presentations have covered AI productivity tools, copyright, privacy, emerging capabilities and the practical implications of using AI in professional work. Before joining King’s, Jason worked as an associate and health economist at Analysis Group and as an economic research analyst at CEP News. His consulting work contributed to published studies of healthcare costs and treatment outcomes. He has also taught economics to CPA students at Sheridan College. Jason’s teaching career has taken him across Canada, Japan and China. In 2024, he was a visiting professor in Canadian Studies at Kwansei Gakuin University in Japan, where he taught Canadian Business Culture and Canadian Economic History. He has also taught Labour Economics at Wuhan University of Technology and Labour Economics and International Trade and Finance at Renmin University in Beijing. His experience teaching in China later prompted him to develop his Economics of China course at King’s. His Canadian teaching experience also includes Wilfrid Laurier University, McMaster University Continuing Education and McGill University. Jason has contributed assessment questions, digital learning activities and instructional materials to economics resources for McGraw-Hill and Pearson. His research examines questions with practical consequences for households, workers and employers. What happens when immigrants cannot find work related to their education? How does a fall in income affect a renter’s ability to move? Do Canadian housing and stock markets share a long-run relationship? His work has appeared in journals including *Applied Economics*, *Regional Science and Urban Economics*, *Cliometrica* and the *Journal of Housing Economics*. As an Associate Researcher with the Montreal Economic Institute, Jason studies how taxes and benefit reductions affect the financial return to work. His policy publications examine the circumstances of low-income seniors, people taking second jobs and workers moving off income assistance. He has also provided media commentary on housing, labour markets, artificial intelligence and the Canadian economy. Jason also holds an MA in Economics from the University of Guelph and an Honours BBA from Wilfrid Laurier University. --- ## MONTREAL ECONOMIC INSTITUTE | RESEARCH REPORTS - **“End Penalties That Punish Low-Income Seniors Who Work”** (2026), with the collaboration of Gabriel Giguère. - **“The Benefits of Eliminating the Middle-Income Tax Bracket”** (2025), with the collaboration of Emmanuelle B. Faubert. - **“Full-Time Struggle and Two-Job Juggle: A Tax Holiday for Canadians to Fight Rising Costs”** (2024), with the collaboration of Renaud Brossard. - **“Canada’s Welfare Wall: Enhancing the CWB for Full-Time Employment”** (2023), with the collaboration of Renaud Brossard. - **“Labour Shortages: Helping a Trapped Pool of Potential Workers Help Themselves”** (2023), with the collaboration of Célia Pinto Moreira. --- ## POLICY COMMENTARY - **“It’s time to give the middle class a real tax break.”** *The Hamilton Spectator*, March 17, 2025. - **“Hey Ottawa, workers with two jobs could use a break.”** *The Hamilton Spectator*, January 13, 2024. - **“Workers Benefit not working for workers.”** *Financial Post*, May 4, 2023. --- ## PEER-REVIEWED RESEARCH | ECONOMICS - Dilmaghani, M., and Dean, J. **“Sexual orientation and intra-household specialization before and after the legal recognition of same-sex marriage in Canada.”** *Applied Economics*, 56(29), 3460–3484 (2024). - Singh, A., McFarlane, A., Dean, J., and Erenburg, G. **“Are Canadian Residential Real Estate and Stock Markets Segmented or Integrated?”** *Empirical Economics Letters*, 23(7) (2024). - Dean, J., and Steele, M. **“Income decline, financial insecurity, landlord screening and renter mobility.”** *Regional Science and Urban Economics*, 95, 103797 (2022). - Dean, J., and Geloso, V. **“The linguistic wage gap in Quebec, 1901 to 1951.”** *Cliometrica*, 16, 615–637 (2022). - Dilmaghani, M., and Dean, J. **“Sexual orientation and homeownership in Canada.”** *Journal of Housing Economics*, 49, 101688 (2020). - Dean, J. **“Does it matter if immigrants work in jobs related to their education?”** *IZA Journal of Development and Migration*, 8, Article 7 (2018). - Dilmaghani, M., and Dean, J. **“Labor Market Attainment of Canadian Jews During the First Two Decades of the 20th Century.”** *Contemporary Jewry*, 38, 49–77 (2018). - Dean, J., and Dilmaghani, M. **“Economic Integration of Pre-WWI Immigrants from the British Isles in the Canadian Labour Market.”** *Journal of International Migration and Integration*, 17(1), 55–76 (2016). - Dilmaghani, M., and Dean, J. **“Religiosity and female labour market attainment in Canada: the Protestant exception.”** *International Journal of Social Economics*, 43(3), 244–262 (2016). --- ## COAUTHORED RESEARCH | HEALTH ECONOMICS AND OUTCOMES - **“Systematic review of long-acting injectables versus oral atypical antipsychotics on hospitalization in schizophrenia.”** *Current Medical Research and Opinion*, 30(8), 1643–1655 (2014). - **“Burden of schizophrenia on selected comorbidity costs.”** *Expert Review of Pharmacoeconomics & Outcomes Research*, 14(2), 259–267 (2014). - **“Impact of omalizumab on emergency-department visits, hospitalizations, and corticosteroid use among patients with uncontrolled asthma.”** *Annals of Allergy, Asthma & Immunology*, 109(1), 59–64 (2012). --- ## OTHER RESEARCH REPORTS - Dean, J., and Dilmaghani, M. **“Religious Diversity and Labour Market Attainment: Nova Scotia and New Brunswick, 1911–2011.”** Atlantic Research Group on Economics of Immigration, Aging and Diversity (2015).