Kota Actions! (New)

Last updated: August 18, 2026

Building your Hyperbound library used to mean working through the bot builder and the scorecard editor one form at a time. Kota Actions changes that: you describe what you need in plain language, Kota drafts it, you review it, and Kota publishes it. Bots and scorecards — creating new ones and updating existing ones — all from a conversation.

This article covers all four Kota Actions:

  1. Create bot(s)

  2. Update bot(s)

  3. Create scorecard(s)

  4. Update scorecard(s)

Why Use Kota Actions?

  • Build at scale. Standing up a library of 50, 90, or 100+ bots for a new team or a new motion

  • Update & Save Time. Ask Kota to change a persona detail, add a scorecard criterion, or adjust weights, and it finds the asset and edits it

  • Stay in one place. Bots, scorecards, and the coaching context behind them all come out of the same conversation.

The more context you give Kota, the less back-and-forth you'll need. Industry, persona, seniority, the tools your buyer uses, the objections you actually hear, all of it makes the draft better on the first pass.

Every Kota Action follows the same four beats.

  1. You describe what you want. Plain language, no template

  2. Kota clarifies. If something is ambiguous (which scorecard? which of the three bots named "Skeptical CFO"?)

  3. Kota builds a draft. Drafts are not live, nothing is visible to your team yet.

  4. You review and approve, then Kota publishes. You'll see a card in the chat with the asset and a confirmation step. Nothing is created or changed until you approve it.

Kota asks for confirmation before any change to an existing asset, not after. If you're mid-conversation and change your mind, just say so, the update won't have happened yet.

1. Create Bot(s)

Ask Kota for the bot you want. Describe the buyer, not the form fields.

"Create a cold call bot: a skeptical VP of Sales at a mid-market SaaS logistics company who already uses Salesloft and doesn't think they need another tool."

For multiple bots at once, say so directly:

Create five discovery bots for our fintech vertical — one each for RevOps Director, CFO, Head of Compliance, VP Engineering, and COO. All mid-market, all mildly interested but budget-constrained.

Then:

  1. Review the draft. Kota shows you the generated persona — name, job title, company, state, goals, objections. Read it for accuracy against how your buyers actually talk.

  2. Request changes in the chat. "Make her more hostile in the first 30 seconds." "She should already have a competing vendor in a pilot." Kota revises the draft in place.

  3. Assign a scorecard. Choose which scorecard reps will be graded against on calls with this bot.

  4. Assign a folder. This controls who can see and use the bot.

  5. Approve to publish. Click Create Agent. The bot goes live and is immediately callable from the AI Roleplay page.

💥 BEST PRACTICE: When creating a batch, create one bot first and get it exactly right. Then ask Kota to "create four more like this one, but for [X]" — you'll carry the quality forward instead of correcting the same thing five times.

2. Update Bot(s)

"Update the Skeptical VP of Sales bot — he should push back harder on pricing and mention that he's been burned by a bad implementation before."

Ask Kota to change an existing bot. You can reference it by name, or describe it.

What happens:

  1. Kota identifies the bot

  2. Kota confirms the change with you

  3. Kota applies the update

💥 TIP: Updates are a good way to keep bots current as your market shifts. When a new competitor shows up in your deals, one Kota request can add that objection across a set of bots.

3. Create Scorecard(s).

"Create a discovery call scorecard for our enterprise AEs. It should cover pain identification, quantifying business impact, multi-threading, and locking in a next step."

Then:

  1. Review the criteria Kota generated

  2. Refine in the chat

  3. Test before you publish. Ask Kota to test the scorecard against a real transcript to see how it grades before it goes live.

  4. Approve to publish.

💥 BEST PRACTICE: Test every scorecard against at least one call you'd consider great and one you'd consider poor. If both come back with similar scores, the criteria need sharpening.

4. Update Scorecard(s)

"Update the Enterprise Discovery scorecard — add a criterion for confirming the compliance review process, and lower the weight on rapport building."

Same flow as updating a bot: Kota identifies the scorecard, confirms the change with you, applies it, and shows you the result. As with creation, you can ask Kota to re-test the updated scorecard against a transcript before you rely on it.

💥 TIP: Scorecard changes affect how future calls are scored. Existing scores are not retroactively recalculated, so it's worth testing an update before publishing it.

Troubleshooting

Kota keeps asking me which bot I mean.

Kota won't modify an asset it can't identify with confidence. Give it the exact bot name, or the folder it lives in.

Kota didn't create anything

Check that you completed the approval step. Drafts stay drafts until you approve them, and a draft that's never approved is never published.

I don't see the create/update options

Kota Actions for assets requires Admin, Owner, or Internal permissions. Members see bot creation only; Observers see neither.

Related Articles

Still have questions? That's what we're here for! Reach out to your Hyperbound support team and we'll help you get the most out of Kota Actions.