Build complex automations through conversation with NIRA, your intelligent copilot for NimbleBrain Studio
NIRA (NimbleBrain Research Assistant) is your AI copilot that transforms how you build automations. Simply describe what you want to accomplish in plain English, and NIRA handles the complexity of configuring tools, orchestrating workflows, and deploying playbooks.
NIRA is an intelligent assistant built into NimbleBrain Studio that understands your business requirements and automatically builds working automations. Instead of manually configuring playbooks, selecting MCP servers, and writing instructions, you have a conversation with NIRA about what you need—and it does the rest.
🗣️ Conversational
Describe your automation needs in natural language—no technical expertise required
🧠 Intelligent
Understands context, selects the right tools, and structures complex multi-step workflows
⚡ Fast
Build automations in minutes that would take hours to configure manually
🔄 Iterative
Refine and adjust automations through continued conversation
Tell NIRA what you want to accomplish in plain English. Be as detailed or high-level as you like—NIRA will ask clarifying questions if needed.Example: “I need a daily report that shows all new support tickets from Zendesk, categorizes them by priority, and posts a summary to our #support Slack channel every morning.”
2
NIRA Analyzes & Designs
NIRA understands your requirements and:
Selects the appropriate MCP servers from the catalog
Designs the workflow structure
Generates playbook instructions
Identifies required API credentials
3
Add Your API Keys
NIRA prompts you for the necessary API keys and credentials. Simply add them when requested—NIRA knows exactly what’s needed.
Your API keys are encrypted and stored securely. They’re never visible to NIRA or logged anywhere.
4
Review the Playbook
NIRA shows you the generated playbook with:
Selected MCP servers
Workflow instructions
Configuration parameters
Expected behavior
You can ask NIRA to make adjustments before deploying.
5
Click GO to Deploy
Once you’re satisfied, click the GO button and NIRA deploys your automation immediately. Your playbook is ready to run!
It’s that simple! Most automations are built and deployed in under 5 minutes.
Request: “Generate a daily sales report from Salesforce showing closed deals, total revenue, and top performers. Send it to #sales-team on Slack every morning at 9am.”What NIRA Does:
Installs and configures Salesforce MCP server
Installs Slack MCP server
Creates playbook to query yesterday’s deals
Formats data into readable summary
Schedules daily execution
Customer Support Automation
Request: “When a new high-priority ticket is created in Zendesk, notify the on-call engineer via PagerDuty and post details in #urgent-support Slack channel.”What NIRA Does:
Sets up Zendesk webhook integration
Configures PagerDuty alerting
Creates Slack notification workflow
Implements priority filtering logic
Inventory Management
Request: “Check weather forecast for our warehouse locations. If severe weather is predicted, send alerts to logistics team and update delivery schedules in our system.”What NIRA Does:
Request: “Combine data from Google Analytics, Stripe, and HubSpot to create a weekly executive dashboard showing traffic, revenue, and conversion rates.”What NIRA Does:
Connects to all three platforms
Aggregates and correlates data
Generates formatted report
Schedules weekly delivery
Anomaly Detection
Request: “Monitor our database for unusual query patterns. If response times spike above 2 seconds, investigate and send a diagnostic report to #engineering.”What NIRA Does:
Request: “When a pull request is opened in GitHub, run our test suite, check code coverage, and post results as a comment. Auto-approve if all checks pass and coverage improves.”What NIRA Does:
Sets up GitHub webhook
Configures test execution
Implements coverage analysis
Creates conditional approval logic
Deployment Automation
Request: “After successful merge to main branch, deploy to staging, run smoke tests, and if everything passes, notify team in Slack with deployment summary.”What NIRA Does:
You: "Create a daily report that pulls sales data from Salesforce and sends it to our #sales-team Slack channel every morning at 9am"NIRA: "I'll create a playbook that:1. Connects to Salesforce using the Salesforce MCP server2. Queries yesterday's closed deals and calculates total revenue3. Formats the data as a summary report with top performers4. Sends to Slack channel #sales-team at 9:00 AM dailyYou'll need to provide:- Salesforce API credentials (Consumer Key and Secret)- Slack workspace tokenReady to add these credentials?"You: "Yes, let's set it up"NIRA: "Great! First, let's add your Salesforce credentials..."
Begin with a basic version of your automation and ask NIRA to add complexity:
“Create a daily sales report from Salesforce”
Once working: “Add revenue trends compared to last week”
Then: “Include top 3 performing sales reps”
Finally: “Send to executives in PDF format”
This iterative approach ensures each piece works before adding more.
Use Business Language, Not Technical Terms
NIRA understands business processes better than technical jargon:Good: “When a customer’s trial ends, send them pricing information”Unnecessarily Technical: “Trigger webhook on trial_end event to execute email_send function with pricing payload”
Describe the 'Why' Along with the 'What'
Providing context helps NIRA build better automations:Better: “We need to track inventory levels because we’ve had stockouts. Alert purchasing when items drop below 20 units so they can reorder in time.”Basic: “Alert when inventory is below 20 units”The context helps NIRA include relevant details in alerts.
Test with Real Data
After NIRA builds your automation, test it:
Run the playbook manually first
Check that outputs match expectations
Verify notifications go to the right places
Ask NIRA to adjust if needed
Leverage NIRA's Knowledge
NIRA knows which MCP servers work well together. Ask for suggestions:
“What’s the best way to automate customer onboarding?”
NIRA currently works best with English, but can understand other major languages. Responses are always in English.
Can NIRA access my data?
No. NIRA only has access to:
Your workspace configuration
Available MCP servers in the catalog
Playbook templates and examples
NIRA never sees your actual data, API responses, or execution results.
What if NIRA doesn't understand my request?
NIRA will ask clarifying questions to better understand your needs. If it still can’t help, it will suggest breaking down the request or provide guidance on how to rephrase.
Can I edit playbooks that NIRA creates?
Yes! NIRA generates standard playbooks that you can view, edit, and customize just like manually-created ones.
Does using NIRA cost extra?
NIRA is included with your NimbleBrain Studio subscription. You only pay for playbook executions and MCP server usage, just like manual playbooks.
Can NIRA help me troubleshoot playbooks?
Absolutely! If a playbook isn’t working as expected, describe the issue to NIRA and it will help diagnose and fix problems.
How does NIRA learn my preferences?
NIRA remembers the context within a conversation session. Over time, patterns in your workspace (like frequently used MCP servers) help NIRA make better suggestions.
When your automation requires multiple steps or data sources, NIRA will suggest creating an Agent:
Example:You: “I need to pull sales data from Salesforce, get market trends from Finnhub, analyze both with AI, and send a report to Slack every Monday morning”NIRA: “This requires multiple data sources working together. I recommend creating an Agent with 4 playbooks:
Salesforce Data - Pull weekly sales data
Market Trends - Fetch relevant market data
AI Analysis - Analyze combined data and generate insights
Slack Report - Format and send to your team
These will run sequentially on Mondays at 9 AM. Would you like me to set that up?”
NIRA handles the complexity of:
Determining which playbooks are needed
Setting up the execution order (sequential or parallel)
Configuring data passing between playbooks
Scheduling the agent to run automatically
Ready for multi-step workflows? Learn how Agents coordinate multiple playbooks → Learn more about Agents