.png)
Thought Leadership
iorad's Interactive Tutorial Modes for Better eLearning
Learn how iorad supports adult eLearning with six tutorial modes, including Try It, View It, Watch It, Print It, Do It, and Quiz Mode.
iorad MCP Turn Your Tutorials Into Instant AI Answers
Connect Claude, ChatGPT, Perplexity, Copilot, or any MCP-compatible AI tool directly to your iorad training documentation. Search tutorials for specific steps, pull analytics insights, and manage your training strategy.
Your process is documented,Now your AI can actually learn it
36%
of L&D teams have setup AI workflows, but just 9% are scaling beyond pilots
Most teams already have AI connected to the process knowledge that would actually make it useful.
52%
of L&D teams name accuracy as a top obstacle to AI adoption
An AI that can't understand each step of your live process isn't accurate, it's guessing with confidence.
72%
of L&D teams expect their biggest AI gains from personalized learning
Personalization at scale needs real process knowledge behind it, not another disconnected AI tool.
*Synthesia: AI in Learning & Development Report 2026
What is iorad MCP?
iorad MCP connects your training context, the tutorials and documented processes you've already built, to the AI assistant your team already uses. Once connected, it can search that content, answer questions from it, track how it's performing, and help you spot what's missing.
Most companies never write any of this information down.
But that's exactly what iorad tutorials are built to capture, and with iorad MCP you can extend that knowledge into every AI conversation.
Share training context with any AI assistant*
Documentation
*Compatible with any MCP equipped tool
Understand how your training performs,Then transform it
Get instant answers from your documentation
Your team already has the answers — they're just buried in tutorials nobody has time to dig through. With iorad MCP, your tutorials become something your AI assistant can search and answer from directly, in seconds.
Understand how your tutorials are performing
Most teams can't say which tutorials are actually working or where learners get stuck without opening a dashboard. That performance data gets surfaced on demand, in plain language, right inside the conversation.
Catch problems before they cost you
A tutorial goes stale, a step quietly trips people up, or access blocks the wrong users, and none of it shows up until it's already a support ticket. One question tells you exactly what needs fixing, no dashboard, no help desk anxiety.
How it works, in three easy steps
Copy & paste one URL, no API keys. Start asking questions, tracking, and identifying improvement areas with ease
https://www.iorad.com/api/mcp
- 1.Add a connectorIn your AI assistant’s settings
- 2.Paste the URL aboveOAuth sign in, no API keys
- 3.Sign in and approve accessThen turn it on inside your chat
Detailed instructions below:
- Open Claude Settings → Connectors
- Select Add custom connector.
- Paste the iorad MCP server URL.
- Approve access in the iorad window that opens (sign in if asked).
- In a conversation, open the Connectors and enable iorad.
- Open ChatGPT.
- Go to Settings → Plugins → Browse Plugins.
- Click +, and enter a name, description, and the URL.
- Approve access in the iorad window that opens (sign in if asked).
- Start a chat by selecting the iorad plugin or using the "@" command.
- Open Perplexity → Connectors → Manage connectors.
- Add a custom MCP connector.
- Enter a Name (iorad), the iorad MCP URL, and other info.
- Approve access in the iorad window that opens (sign in if asked).
- In a conversation, turn on iorad from the + menu, or type @iorad.
Prompts from the team
Try one of these example prompts to see how it works. Swap anything in [brackets].
What you get: A ranked list by views and completion.
Prompt: Using the iorad MCP, pull analytics for every tutorial in [library name]. Rank the top 10 by a mix of views and completion rate, not views alone. Flag any tutorial that over-indexes on views but under-performs on completion.
What you get: Three plain-language sentences, no dashboard needed.
Prompt: Using the iorad MCP, pull tutorial and library analytics for [time period]. Summarize performance in exactly 3 sentences a non-technical exec would understand: overall usage, the standout tutorial, and where attention is needed.
What you get: A clear before/after instead of two reports to compare by eye.
Prompt: Using the iorad MCP, compare tutorial analytics for [this period] against [last period]. Call out anything that moved more than 20%. Ignore anything under 5% — that's noise.
What you get: A ready-to-edit draft — title, intro, numbered steps.
Prompt: Using the iorad MCP, pull the tutorial titled [tutorial name] and turn it into a help-center article: a title, a 1-sentence intro, and the steps as a numbered list. Stay accurate to what the tutorial shows.
What you get: The exact steps from the matching tutorial, summarized in plain language, with a link back.
Prompt: Using the iorad MCP, find the tutorial that covers [process] and answer my question: [your question]. Base your answer only on what's in the tutorial. If more than one tutorial covers this, use the most recent one and tell me which you chose. Include a link.
What you get: A short, ranked list of matches with what each one covers.
Prompt: Using the iorad MCP, search my tutorial library for anything related to [topic]. List up to 5 matches with the tutorial name, which library it's in, and one line on what it covers.
What you get: Every tutorial in the library, ordered for someone brand new.
Prompt: Using the iorad MCP, list every tutorial in the [library name] library. Group them into a suggested order for a new hire, with a one-line reason for the order. Format as a checklist.
What you get: A condensed step-by-step summary short enough for Slack or email.
Prompt: Using the iorad MCP, pull the tutorial titled [tutorial name] and summarize it in under 150 words, keeping every step but in plain sentences. End with a link to the full tutorial.
What you get: What's different, which is more current, and whether one should be retired.
Prompt: Using the iorad MCP, compare [tutorial A] and [tutorial B]. Tell me what's different, which is more recently updated, and whether one should be retired or merged.
What you get: Tutorials grouped by topic with a one-sentence summary per group.
Prompt: Using the iorad MCP, list everything in the [library name] library. Group the tutorials by topic with a one-sentence summary of what each group covers.
What you get: A chronological feed of who shared what, and when.
Prompt: Using the iorad MCP, show me everything shared or published across the team in the last [30] days. List who shared it, what it was, and when.
What you get: A list ready to send as update reminders.
Prompt: Using the iorad MCP, find tutorials in [library name] untouched for over [90] days, grouped by original creator, with recent performance for each.
What you get: Creators ranked by tutorials created/updated, with their most recent one.
Prompt: Using the iorad MCP, pull creator activity for [time period]. Rank team members by tutorials created or updated, with each person's most recent tutorial.
What you get: Tutorials created/updated paired with how well they actually perform, delivered as a drafted email.
Prompt: Using the iorad MCP, pull tutorial output and performance for [team member, or 'the whole team' to rank everyone] over [time period] — what was created or updated, plus launches and completion rate since. If it's one person, keep it factual for a 1:1; if it's the whole team, rank by a mix of volume and how well the work performs, not raw count. Draft it as an email to [manager email], ready for me to review. (Requires a Gmail connector alongside iorad.)
What you get: Creator and last editor, with dates.
Prompt: Using the iorad MCP, tell me who created [tutorial name] and who last edited it, and when.
What you get: The step, what it asks, and how many learners got it wrong, ranked.
Prompt: Using the iorad MCP, pull the quiz results for [tutorial name]. Find the steps with the most incorrect answers, tell me what each step asks, and rank them by how many learners got it wrong.
What you get: A shortlist with high views but low completion.
Prompt: Using the iorad MCP, pull analytics for [library name]. Find tutorials with above-average views but below-average completion, ranked by the size of the gap.
What you get: Who deactivated what, and whether it still mattered.
Prompt: Using the iorad MCP, find tutorials deactivated in the last [90] days. For each, tell me who deactivated it and when, and flag any that still had meaningful views or completions before being turned off.
What you get: The step, what it asks the learner to do, and how many people exited there.
Prompt: Using the iorad MCP, pull the analytics for [tutorial name]. Find the step with the most learner exits across all playback modes, then pull the step content to tell me what it asks the learner to do. List the top 3 exit points, ranked by exit count.
What you get: Settings and privacy status, duplicate or near-duplicate tutorials flagged, and real learner performance — plus a verdict on whether it's ready for real use.
Prompt: Using the iorad MCP, give me a complete health check on my [library name] library. Pull its settings and privacy status, list every tutorial, flag any duplicate or near-duplicate titles, and pull overall learner performance — launches, completions, unique learners. Tell me whether this library looks ready for real use or still needs cleanup, and why.
What you get: A task with the exact step, what it's asking learners to do, and how many dropped off — ready to hand to whoever owns that tutorial.
Prompt: Using the iorad MCP, find the tutorial in [library name] with the most learner drop-off, identify the top step where people exit or answer incorrectly, and pull what that step asks the learner to do. Then create an Asana task titled '[tutorial name] — fix step [X]' in [project name], assigned to [person], with the step details and exit count. (Requires an Asana connector + iorad.)
What you get: Every tutorial whose privacy setting is likely blocking real visitors, logged as a row in your Notion tracker with the owner tagged so it gets triaged, not lost in a Slack thread.
Prompt: Using the iorad MCP, list our tutorials by privacy setting, Public, Unlisted, Embed Only, Invite Only, Locked, and flag any set to Invite Only or Locked that are also embedded on a page. Look up who owns each flagged tutorial, then add a row to my [Notion database] with the tutorial, the owner, and why it's likely blocking visitors. (Requires a Notion connector + iorad.)
What you get: A ranked weekly digest delivered straight to a channel, not just a report you copy over yourself.
Prompt: Using the iorad MCP, build this week's tutorial performance report: pull the last 7 days vs. the previous 7, top 5 by views with completion rate and week-over-week change, and the biggest mover. Then post it to [#channel] in Slack. (Requires a Slack connector alongside iorad.)
What you get: A per-site performance breakdown for your most-used tutorials, so weak spots hiding at specific embed locations don't stay hidden.
Prompt: Using the iorad MCP, for our top 10 tutorials, list every website or page where each one is embedded, along with performance at that location. Flag any embed location where a tutorial gets traffic but its completion rate lags well behind how it performs everywhere else.
Explore More Features
Related content
See all
Comparisons
iorad vs. Arcade
Compare iorad and Tango for enablement, documentation, workflow training, integrations, and scale to choose the right tool for your team.
Comparisons
iorad vs. Scribe
Compare iorad and Scribe for workflow documentation, training, updates, interactivity, pricing, and scale to find the right tool for your team.
Comparisons
iorad vs. Mindtickle
iorad vs. Mindtickle compared: sales readiness and coaching vs. interactive software training, plus pricing and which tool fits your team.
Want to learn more about iorad MCP?
© Copyright 2026 – iorad Inc. All rights reserved.
