What Interviewers Actually Probe
Teams development interviews rarely ask you to recite manifest schema field names from memory; they probe whether you understand why the platform is built the way it is — why tabs run in an iframe, why bots need a public messaging endpoint, why SSO requires a server-side token exchange — because that understanding is what lets you debug novel problems on the job. Expect a mix of conceptual questions (explain the difference between a tab and a bot), applied questions (write code for a specific TeamsJS or Bot Framework scenario), and scenario questions (a customer reports X, how do you investigate) in roughly equal measure.
Cricket analogy: It's like a talent scout not just asking a bowler their economy rate but watching how they adjust their line when a batter starts stepping out — interviewers care less about memorized manifest fields and more about whether you understand why the platform behaves the way it does.
Architecture and SDK Questions
Be ready to clearly articulate the difference between the four core surfaces: a tab is an iframe-hosted web app rendered inside Teams for persistent UI, a bot is a conversational endpoint built on the Bot Framework that responds to messages and can send proactive notifications, a message extension lets users search or act on external data from the compose box or message context menu, and a connector or webhook posts notifications into a channel from an external system. Also expect questions on the TeamsJS SDK's capability-checking pattern — calling capability-specific isSupported() checks (like dialog.isSupported()) before invoking a feature, since not every Teams host (desktop, web, mobile, Outlook) supports every capability.
Cricket analogy: It's like distinguishing a specialist batter, a strike bowler, an all-rounder, and a wicketkeeper — each role (tab, bot, message extension, connector) has a distinct job on the team, and confusing their responsibilities is a rookie mistake in an interview.
Authentication and Graph API Questions
A frequent question is to explain the On-Behalf-Of flow end to end: the tab calls authentication.getAuthToken() to get a Teams SSO token scoped to your Azure AD app, sends that token to your backend, the backend exchanges it for a Graph access token via MSAL's acquireTokenOnBehalfOf using the app's client secret or certificate, and only the backend ever calls Graph with the resulting token. A related question tests whether you know the difference between delegated permissions (the app acts as the signed-in user and is bound by that user's own access rights) and application permissions (the app acts with its own identity and typically requires admin consent, appropriate for background services with no signed-in user context).
Cricket analogy: It's like a team manager (the backend) presenting a player's official accreditation (the SSO token) to match officials to get a fresh access pass (the Graph token) rather than the player trying to negotiate access directly — the OBO flow always routes through this intermediary.
Scenario-Based Debugging Questions
A classic scenario question: 'A tab works fine on desktop but fails to load only on Teams mobile — how do you debug it?' A strong answer walks through checking whether the failing capability is even supported on mobile (via isSupported() checks), whether the tab's CSS assumes desktop viewport dimensions, and whether the issue is a Content-Security-Policy or cross-origin restriction that behaves differently in the mobile WebView versus desktop's Edge WebView2 host. Another common scenario is 'users report a bot stopped responding after working fine for months' — a strong answer investigates whether a bot's stored conversation reference expired or became invalid, whether an Azure AD app secret recently expired (a very common real-world cause), and whether Bot Framework Connector service health has any reported incidents, using Application Insights correlation IDs to narrow it down rather than guessing.
Cricket analogy: It's like diagnosing why a bowler who was effective all season suddenly can't get wickets on a specific pitch — you check conditions specific to that pitch (mobile WebView) rather than assuming the bowler (the tab code) itself changed.
// A strong interview answer often includes a defensive capability check like this
import { dialog } from '@microsoft/teams-js';
async function openTaskDialog() {
if (!dialog.isSupported()) {
// Fall back gracefully on hosts (e.g. some mobile contexts) that don't support dialogs
console.warn('Dialog API not supported on this host; falling back to inline UI');
return;
}
dialog.open({
title: 'Create Task',
url: 'https://contoso.example/task-dialog',
size: { height: 400, width: 500 },
});
}For scenario questions, structure your answer like a STAR response but technical: state your hypothesis, name the specific tool you'd use to confirm it (Application Insights, DevTools extension, Bot Framework Emulator), then state the fix — interviewers reward a clear investigation process over guessing the right answer immediately.
A common pitfall in interviews is confusing proactive bot messaging (sending a message without a prior user turn, which requires a saved conversation reference) with reactive messaging (replying within an existing turn) — mixing these up when explaining bot architecture is an easy way to lose credibility.
- Interviewers probe understanding of why the platform works a certain way, not memorized manifest schema fields.
- Be precise distinguishing tabs, bots, message extensions, and connectors — each has a distinct, non-overlapping role.
- Know the OBO flow end to end and the difference between delegated and application Graph permissions.
- Practice capability-checking with isSupported() before invoking TeamsJS features that vary by host.
- For scenario questions, walk through a clear investigation process rather than guessing the answer immediately.
- Expired Azure AD app secrets are a very common real-world cause of 'the bot suddenly stopped working.'
- Don't confuse proactive messaging (needs a saved conversation reference) with reactive in-turn replies.
Practice what you learned
1. What distinguishes a Teams bot from a Teams tab architecturally?
2. What is the key difference between delegated and application Graph permissions?
3. Why should a TeamsJS feature call isSupported() before invoking a capability like dialog.open()?
4. A bot that worked for months suddenly stops responding. What is a very common real-world cause worth checking first?
5. What is the difference between proactive and reactive bot messaging?
Was this page helpful?
You May Also Like
Teams App Best Practices
Practical guidelines for building performant, secure, and visually consistent Microsoft Teams applications that hold up in production.
Microsoft Teams Quick Reference
A condensed cheat sheet of core manifest fields, TeamsJS APIs, bot patterns, and Graph endpoints for Teams app development.
Testing and Debugging Teams Apps
Practical techniques for local debugging, testing SSO and bots, and diagnosing issues in production Teams apps.
Related Reading
Related Study Notes in Microsoft Technologies
Browse all study notesWindows 10 / UWP Development Study Notes
.NET · 30 topics
Microsoft TechnologiesWindows Batch Scripting Study Notes
Batch · 30 topics
Microsoft TechnologiesMFC (Microsoft Foundation Classes) Study Notes
C++ · 30 topics
Microsoft TechnologiesSilverlight Study Notes
.NET · 30 topics
Microsoft TechnologiesXAML Study Notes
.NET · 30 topics
Microsoft TechnologiesWPF (Windows Presentation Foundation) Study Notes
.NET · 30 topics