The Dapti welcome screen, with quick actions, a chat input, connected tools, and your recent conversations.
The Dapti welcome screen.
Dapti reasoning through a question.
A complete Dapti response with tables, links, an interactive button, and message actions.
Message actions appear under every Dapti response.
The Dapti home page.
The conversation top bar.
The conversation history popover.
The Activity panel showing Dapti's reasoning.
Message actions appear under every response.
The feedback panel that appears when you click thumbs down.
Dapti gathers context, builds the agent, and opens a side preview with the full configuration.
Dapti diagnoses the weak section, proposes a stronger version, and asks for confirmation before applying.
Dapti shows the current settings, proposes new values, and waits for you to confirm.
Dapti picks the template, collects context, and opens a live preview of the agent it just built.
Dapti diagnoses what is missing and proposes a structured replacement.
Dapti turns natural-language behavior requests into structured prompt changes.
Dapti returns a structured metrics table and a written analysis of what stands out.
Dapti pairs the structured outcome data with a human-readable summary of each conversation.
Dapti finds the structural issues across calls and proposes specific, prioritized improvements.
Dapti generates, deploys, and wires up the flow. The visual canvas opens on the right so you can see what was built.
Dapti groups flows by folder and tells you which ones have run, which have not, and what to try next.
Dapti rebuilds the flow with the new step, runs a smoke test, and reports exactly what changed.
Dapti's full health check: a phased scorecard, an Account Context block, and a plain-English summary of where you stand.
Dapti diagnoses the blocking issue and returns a prioritized fix list, not a generic checklist.
Dapti picks the single highest-impact next step, explains why, and gives you exactly the steps to do it.
Dapti returns a side-by-side comparison and explains what each metric difference actually means in your industry.
Dapti turns peer behavior into a prioritized action list specific to your account.
Dapti translates the benchmark gap into concrete prompt changes and surfaces a relevant Academy tutorial video.
Every meeting the Note Taker recorded, in one clickable list.
Dapti's structured meeting brief: overview, topics, stakeholders, action items, AI sales tips, and call insights.
Dapti clusters questions and objections across multiple meetings and translates them into strategic recommendations.
Dapti's credit overview: plan, balance, usage percentage, and contextual guidance based on where you are.
Dapta plans side by side, plus a recap of where you stand today.
Dapti runs the math and tells you whether your current plan can cover the workload.
Dapti plans the full setup and stops to ask for any missing integrations before it builds.
An empty-state response that also gives you four starting points to set up your first schedule.
Dapti turns "what should I automate?" into three concrete, business-focused recommendations, each with the exact prompt to activate it.
A live web search, summarized in plain English, with a stats table, cited sources, and a downloadable report.
Dapti reads the page, answers your question, and offers a downloadable summary.
A multi-source best-practices guide with an executive summary, 8 structured sections, and a downloadable .docx.
Dapti points you to the Plans page and to sales if you need a tailored conversation.
Dapti gives you the fastest path to a human when the request goes beyond what it can answer directly.
Dapti returns the program structure, your reward, payout timing, where to find your referral link, and who to contact for questions.
 Then in the menu click on Profile, there you will find a link for “Refer & Earn 50.00”\ \  * Your code will appear; you can copy it and share it with your referrals. Further down, you'll also be able to see:\ \ \- Your total number of invitations\ \- Your potential earnings\ \- Your earnings with Dapta
 \ **Step 2: Share your Referral Code** \ Your invitations become effective when your referral makes a payment. How is the payment made?\ Here's how:\ Your referral must follow these steps: Go to the bottom left corner,\ click on your email address, and select the Billing option, and within that, the Billing tab  * On the “Billing” tab, you will see “Enter your referral code”, enter the code and click on **“Redeem code”.**  * When you see the green message **“Your referral code successfully claimed”,**\ you can now make the payment for one of Dapta's plans by clicking on the **“Start Now”** button.  * It takes you to the Stripe screen and you make the payment. **That's it!** You will automatically be charged an additional 100,000 credits that month, on top of your plan's allowance. **Step 3: Check which of your referrals made the payment** * To see who made the payment, click on the “Refer & Earn” banner,\ and you will see the people who redeemed your code and are:\ \- **Pending:** They have not yet made the payment\ \- **Earning:** They did make the payment ☝️ **Remember:** We will deposit your $50.00 reward in cash (or Dapta credits, if you prefer) during the first week of the month following your referral's payment.\ **For example,**\ If your referral made a payment on October 12th,\ we will process your payment during the first week of November. Any questions or inquiries can be directed to:
# How to create an account in Dapta? In this guide we show you step by step how to create a Dapta account. {% embed url="
*** ### **🌟** Improve with AI (Below Prompt) Use this option to improve your entire prompt at once.\ \ 💡 How It Works: 1. 📄 Scroll to the bottom of the prompt editor. 2. ✨ Click the **“Improve with AI”** button. 3. 🤖 The AI will analyze your full prompt and suggest broader changes—this might include rewording, restructuring, or improving tone.
💡 **Why it matters:** Your feedback helps us train the AI to get smarter and more accurate over time. Every comment helps improve future calls. # Post Call Analysis – Call Analysis Flow 🔀 This section allows you to configure how calls made by the AI agent will be analyzed using the flows previously created in Flow Studio. #### Select Call Analysis Flow 
Click on “Select Call Analysis Flow” 
You will only see flows that meet the specific criteria for call analysis. If your flow doesn’t appear in the list, it’s because it doesn’t meet the following conditions: * The flow must have a “Trigger” type that receives all parameters through the body of the request. * The flow must respond with text, since LLM models can only process text-based responses. * Only one analysis flow can be selected in this feature.
Once the flow is selected, you’ll see the flow’s URL and API Key already filled in under the “Call Analysis Flow” field.
You can extract the following data types from post-call analysis: #### Text (String): Used to capture written, detailed information such as names, descriptions, and open-ended responses.
Example: Transaction amounts, call duration, or satisfaction scores. # Post Call Analysis – Notifications 🔔 Automatically send an email alert after a call, choosing who receives it, what call data it includes, and the post-call variable condition that triggers it. The **Notifications** feature lets your agent send an automated email after a call is executed. You decide who receives it, which post-call data the email contains, and the condition that must be met for the email to be sent. Emails are sent from `notifications@daptatech.com`. You'll find it in the **Post-Call Analysis** section of your agent, selectable from the **Call Analysis** tab within the voice agents, below **Call Analysis Flow**.
*Minimum: 0 (Rarely interrupted)* *Maximum: 1 (Highly interruptible)* *Recommended: 0.5–0.8 for balanced flow›*
Fine-tuning this gives users more conversational control—without making the AI feel passive. # Speech Settings - Active Listening 🎧 Enable subtle backchanneling cues—like “yeah,” “uh-huh,” or light affirmations—to make your AI sound more present and engaged during human speech.
#### The available options are: * Café: Simulates the sound of a coffee shop * Convention Center: Simulates the sound of a conference or convention * Outdoors – Summer: Simulates the sound of a summer outdoor environment * Outdoors – Mountain: Simulates the sound of a mountain outdoor environment * Static Noise: Simulates static background noise * Call Center: Simulates the sound of a contact center * None: No background sound is added **Ambient Sound Volume**
To address this, Dapta provides the “Boosted Keywords” functionality, which allows you to: * Customize the system’s vocabulary according to specific needs * Improve accuracy in your agent’s specialized domains
will be transformed into: > “Call my number two one three seven one one two three four two on July fifth, twenty twenty four for the twenty four dollars twelve cents payment” It’s important to note that this feature adds a small latency (approximately 100 ms) to the overall process. *** #### Language Configuration Currently, speech normalization is supported in the following languages: * English * Spanish * French * German For other languages, this feature will not make any modifications to the text. If you select a non-multilingual language, the normalization will use that language’s rules (for example, “1” will be normalized to “one” if English is used). If you select the multilingual option, the system will automatically detect the appropriate language based on the generated text and normalize it accordingly. # Configure Voice – Voice Model 🤖 Choose the voice model your AI agent will use
# Configure Voice – Voice 🗣️ Choose your agent’s voice
To use the voice you’ve selected, simply click ✔️ Use Voice. # Voice Settings - Voice Speed 🏃➡️ Adjust how quickly the AI speaks—faster for efficiency, slower for clarity. This setting fine-tunes pacing to match your use case, from rapid-fire support calls to thoughtful, deliberate conversations.
*Minimum: 0.5* *Maximum: 2.0* *Recommended: 0.7–1.3 for balanced flow and comprehension* # Voice Settings - Voice Temperature 🌡️ Voice Temperature controls the **creativity** and **variation** in the AI’s speaking style. Higher values introduce more unpredictability in tone, pacing, and emphasis—making the voice sound livelier or more expressive. Lower values make it more consistent, focused, and neutral.
*Minimum: 0* *Maximum: 2* *Recommended: \~1 for default behavior* # Voice Settings - Voice Volume 🎚️ Voice Volume sets how loud or soft the AI sounds during a call. This controls gain—not energy or tone—so you can match the AI’s voice level to different environments, hardware setups, or user preferences.
*Minimum: 0* *Maximum: 2* *Recommended: 1 for standard volume balance* # Conversation Settings - Voicemail Detection 📬 Voicemail Detection is mission-critical for ensuring your agent behaves correctly when a user doesn't pick up. Basically, this tool stops your agent from having a deep conversation with a recording—which acts as a great fail-safe.\ \ We highly recommend keeping this toggled ON to ensure maximum call efficiency. To access it, go to your voice agent and hit Settings > Call Settings.
This allows you to ensure that if a conversation is not productive but no silence is detected, the call will still end at the set maximum duration — helping avoid unnecessary credit usage. # Verifying an Existing Phone Number for Outbound Calls {% embed url="
Use the Phone Numbers page to verify, purchase, and manage numbers for Voice Agents.
Use the Sip Trunks page to add and manage SIP trunk extensions for Voice Agents.
The Emotionality rating in the Select Voice window.
Emotion tags shown inline on the agent's lines in the transcript.
**🔧 Workflow Overview** 1. **Connect your HubSpot account** 2. **Fetch a list of contacts from HubSpot** 3. **Trigger calls using your AI voice agent** 4. **Capture call data (recording, transcript, status)** 5. **Log results and update contact records in HubSpot** Let’s start with Step 1: **Connecting HubSpot credentials and configuring the “Get List” node** 👉\ \ 🔑 **Step 1: Connect Your HubSpot Account** To get started, you'll need to authorize your HubSpot account so Dapta can pull contact data and update records. **🧭 How to Connect HubSpot:** 1. Click your profile avatar in the bottom-left corner of Dapta.\ \ 
2. Select **Credentials** from the dropdown. 3. Click the **➕** button next to HubSpot.
There are two main trigger types: * **Webhook (HTTP Request)** – Starts the flow when an external service sends a request to this URL. Useful for syncing data from CRMs like HubSpot or responding to user actions. * **Schedule** – Runs the flow automatically on a schedule (e.g., daily, hourly). For this use case, we’re using a **Webhook Trigger** with a `GET` method so the flow activates when a GET request is sent to the provided Webhook URL. [🔗 Learn more about Trigger types](/flow-studio/setting-up-triggers-in-flow-studio)\ \ 
Now that we’re looping through each contact, it’s time to place the calls. **Search & Add Node**\ In the Actions panel, search for `Dapta Phone Call`.\ \ \ \ Drag it into the canvas and connect it to the Loop List node. **Configure the Call**\ Click into the node to set your call parameters: * Choose your **Voice Agent** * You can map the variables like `{{item.firstname}}`, `{{item.phone}}`, etc. * Adjust call speed and retry logic if needed\ [🔗 Learn more about Voice Agent Settings](/ai-voice-agents/how-to-set-up-your-ai-voice-agent/voice-agent-settings-guide)\ \ 
This node will dial each contact in your HubSpot list using the selected voice agent, one at a time. 📤 **Step 4: Configure Post‑Call Data Retrieval**\ \ [🔗 Learn Extraction Information From Your Calls](/ai-voice-agents/extracting-information-from-your-calls)\ \ **1. Go to Your Agent Settings**\ Open your agent in Dapta, then navigate to the **Call Analysis** tab.\ \ **2. Select a Call Analysis Flow**\ Click **“Select Call Analysis Flow”** and choose a Dapta flow you’ve created.\ \ \ \ **3. Optional Add Post‑Call Fields**\ Under **Post Call Data Retrieval**, click **“+ Add”** to define the insights you want extracted.
* In the **Request Body** panel, click **“Copy to input.”**\ \ \
**View Request Body** * Expand the **Request Body** panel to inspect exactly what data was sent. \ 
3. **Trigger the Logging Flow** * Hit **“Run Flow”** 4. **Confirm CRM Logging** * Check HubSpot Contacts to verify the call was logged correctly.\ \ ✅ **Once the flow runs successfully, the call will be logged inside the contact’s activity timeline in HubSpot.**\ \ 
# Designing Voice Agent Prompts for iOS Call Screening ### Overview Apple’s iOS 26 introduces a system-level call screening feature that automatically answers calls from unknown numbers and asks the caller to explain who they are and why they are calling. Because a large portion of outbound calls reach iPhone users, this change directly impacts how voice AI agents should be designed and deployed. Dapta does not bypass or disable iOS Call Screening. Instead, success depends on how your **agent prompt, opening message, and call behavior** are configured. This guide explains how iOS Call Screening works and provides con
* **Discover Sitemap:**
* Type or paste text directly. * Click Save.
* **Add an Example Question Manually**