“AI as Co-Pilot: Human-Commanded Automation for Smarter Decisions”Record inserted or updated successfully.
AI & Data Management

“AI as Co-Pilot: Human-Commanded Automation for Smarter Decisions”

Author : CA. Preksha Jain

Watch on Youtube

The Challenge Every CA Faces

As Chartered Accountants, we live in a paradox.

  1. On one hand, our profession is built on trust, confidentiality, and compliance.
  2. On the other hand, growth requires speed, scale, and constant engagement with clients and prospects.

Traditional CRMs help us organize, but they fall short in two ways:

  1. They rely on manual updates, which drain time and create blind spots.
  2. They encourage push marketing tactics, like cold mails, which are not ethical—or even permitted—for us as CAs.

This gap between compliance and client engagement is what inspired me to build a new kind of CRM.


The Vision: A Smarter, Ethical CRM

My goal was simple but ambitious:

  1. Automate the mundane.
  2. Respect compliance boundaries.
  3. Create growth through pull, not push.

This is not just about building another dashboard. It’s about reimagining what a CA-focused CRM system should look like in the age of AI.


The Architecture I Designed

1. Frontend: Lovable CRM Dashboard

I started with Lovable, which allowed me to design a real-time dashboard directly connected to a Google Sheet containing CRM data.

  1. Every lead, every client record, every interaction is instantly visible.
  2. An AI CRM Assistant sits on top of this dashboard, acting as a human-like interface. You don’t “use” the CRM anymore—you talk to it.

Want to send an email to a client? Ask the assistant.

Need to update client details? Tell the assistant.

This turns CRM into a conversation, not a chore.






Prompt For CRM Dashboard


Create a visually engaging, beautifully designed **dark theme CRM dashboard** for a Chartered Accountant firm that integrates with my Google Sheets database in real time. The dashboard should automatically refresh on load and display comprehensive client and business analytics, including:


- Revenue trends by service line (Tax, Audit, GST, Advisory)

- Lead conversion funnel (New → Proposal → Won → Lost)

- Partner/Manager performance metrics

- Weighted pipeline revenue and realization %

- Client health indicators (churn risk, feedback scores)

- Source effectiveness (referrals, LinkedIn, events)

- Geo-distribution of clients


Here is my google sheet link for the data intergration : https://docs.google.com/spreadsheets/d/1qIj-cu8/export?Format=csv

Design specifications:

- **Modern glassmorphism effects** with subtle neon accents (blue, teal, purple) for a fintech aesthetic

- **Smooth animations** and transitions between charts and widgets

- **Interactive data visualizations**: bar charts, funnel charts, line graphs, heat maps

- **Responsive layout** with customizable widgets that users can rearrange

- **Professional finance look**: clean typography, clear numeric emphasis, visually appealing but not distracting

- Hover effects and gradient backgrounds that feel dynamic, like a high-tech financial control panel

- KPI cards at the top, charts in the middle, detailed tables below

- Sleek, futuristic design that still feels appropriate for a CA practice, like a modern Bloomberg terminal

The end result should make someone say “wow” while still being **practical, trustworthy, and professional** for a Chartered Accountant firm.


  1. Backend: n8n + AI Agent
  2. The real magic happens behind the curtain.
  3. The assistant is wired to n8n webhooks, which trigger an AI agent.
  4. This agent doesn’t just process commands—it executes them.
  5. Updating Google Sheets
  6. Sending emails
  7. Adding or deleting client records
  8. Logging interactions

In short, the assistant doesn’t just “promise”—it delivers.





Prompt For My AI Agent in N8N for CRM Dashboard


You are an AI Agent that manages information in a Google sheet. Your job is to read incoming chat messages, review the current sheet data and update any relevant information as needed.

Core Process

1. Read the message: Understand what the user is asking for or reporting.

2. Review the sheet: Use the read tool to get current data from all relevant sections.

3. Analyze: Determine what information needs to be updated, added, or modified.

4. Update: Use the update tool to make necessary changes.

5. Confirm: Briefly explain what you updated.

Available Tools

A) read_google_sheet : Retrieves current data from the Google sheet.

B) update_google_sheet: Updates specific cells or ranges in the sheet.

Guidelines

a) Always read the sheet first before making any updates.

b) Only update information that has actually changed or is newly provided.

c) Be precise with your updates – don’t modify unrelated data.

d) If the message is unclear, ask for clarification before updating.

e) Provide a brief summary of what you updated after making changes.

Response Format

1. Read the current sheet data.

2. Identify what needs updating based on the message.

3. Make the necessary updates.

4. Confirm: “Updated (specific information) in the sheet.”


Stay focused, efficient, and accurate in your updates.

Whatever the message you will be sending will be from the side of CA Preksha Jain.




Prompt for AI Assistant in Lovable for CRM Dashboard

               I want you to give me a chat bot besides this , the chatbot should go to the web hook:

https://app.n8n.cloud/webhook-test/6c9

it should get the response from the web hook and automatically update over here


3. Lead Generation Workflow: Ethical by Design

Here’s where the design philosophy really shows.





  1. A workflow in n8n scrapes leads from Google Maps, updating their details in the CRM sheet automatically.
  2. From the lead’s website, it extracts valid email IDs.
  3. Instead of blasting cold mails, it sends a permission mail—asking if they’d like to opt-in for newsletters or updates.

This changes everything.

  1. No cold outreach.
  2. No compliance risk.
  3. Only interested, engaged leads enter the funnel.

We don’t push. We pull. And that makes all the difference.


API Key for the Google Map Scraping for lead Generation

Rapidapi.com – categories – data- Google maps scraper-Api for places

Code for Extracting URL from websites for further extraction of mail id from it.


// Get all rows from Google Sheet

const items = $input.all();


// Regex for extracting URLs

const regex = /(https?:\/\/[^\s"'>]+)/g;


let results = [];


for (const item of items) {

 // Get website column (change "Website" to your actual column name)

 const input = item.json.Website || "";

 

 // Extract all URLs from this row

 const websites = input.match(regex) || [];

 

 // Push each website as its own output item

 for (const website of websites) {

   results.push({ json: { website } });

 }

}


return results;

Code for Extracting Mail ID from Websites

const input = $input.first().json.data

const regex = /[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.(?!jpeg|jpg|png|gif|webp|svg)[a-zA-Z]{2,}/g

const emails = input.match(regex)

return{json: {emails:emails}}



Content of Mail for sending to the extracted mail ids

         Subject : Permission to send practical guides on GST compliance and small-business tax savings.

            Hello,

I’m CA Preksha Jain. We prepare short, practical guides on GST compliance and small-business tax savings.


May I send one short guide (no obligation) that could be helpful? If yes, I’ll send it to this address; if not, no problem — I won’t follow up.


Regards,

             CA Preksha Jain



Why This Matters for CA Firms

ProblemTraditional CRMAI-Powered Ethical CRM

Data UpdatesManual, delayedAutomated, instant
Client CommunicationGeneric, pushPersonalized, pull
Compliance RiskHigh with cold outreachZero—consent-based
ProductivityLost in admin workFreed up for advisory
GrowthReactiveProactive & scalable

This system doesn’t just save time. It shifts the role of technology in a CA firm—from being a passive tool to being an active compliance ally and growth driver.


The Bigger Picture: Future of AI in CA Practice

This project is more than a CRM experiment. It’s a proof of concept for what’s possible when we combine:

  1. AI agents for execution,
  2. No-code platforms like n8n for workflows, and
  3. Low-code frontends like Lovable for experience.

The future CA firm won’t be a place where juniors update spreadsheets late into the night.

It will be a place where AI takes care of the mechanics, while we, as professionals, focus on strategy, insights, and human trust.


Closing Thought

What I’ve built is not just a CRM. It’s a blueprint for a new way of working—one that balances compliance with innovation, and efficiency with ethics.

And perhaps the most important lesson?

Growth doesn’t come from pushing harder.

Growth comes when you design systems that make clients want to come to you.