AI & Digital Transformation
TallyGPT: Transforming Accounting Systems with Smarter AI- Finance Tools
Watch on Youtube
Slide 1: Title
- Subtitle: Analytics & Create Rapid Tally TDL with Zero Coding Skills Needed
- Presenter: CA Ankush Jain, FCA | CISA (US) | FAFD | DISA | CCAB(ICAI) | B.COM.(H) | CEH(US) | ISO 27001 LA | PCI DSS Imp (TUV) | Full Stack Software Developer |
- Slide 2: Disclaimer
- This presentation is for informational purposes only and is based on personal research and experience. It is not intended to promote or demotivate any technology or application.
- The implementation of any technology solutions should be evaluated against your organization's specific requirements, security policies, and regulatory obligations.
- Application of these use cases should be done in a safe environment and with proper guidance and research only.
Slide 3: The Challenge
- Traditional Tally TDL Development Challenges:
- Steep learning curve with complex syntax and definitions
- Time-consuming development process
- High risk of syntax errors and debugging difficulties
- Limited access to comprehensive documentation
- Difficulty integrating advanced features like ODBC, external files, etc.
Slide 4: The Solution - Local LLM Deployment
Let’s Connect Tally Software with AI Tools for Data Analytics & Rapid TDL Creations.

Tally Data Analytics and AI-Enhanced TDL Development Pipeline.
✓ Prompt Engineering - Converting user requirements to expert-level prompts
✓ Knowledge Acquisition - Systematic collection of TDL documentation and examples
✓ Knowledge Processing - Creating optimized standalone reference document
✓ AI Integration - Configuring Claude with MCP and with specialized TDL knowledge
✓ ODBC Connection - Direct database access for real-time analytics and TDL generation
Slide 5: Demonstration Overview & Tech Stack Used for Implementation
1. ChatGPT: For Prompt Engineering.
2. Claude Desktop: AI-Tool for connecting Tally with AI and using NLP Chat prompt and Coding TDL.
3. MCP Tools:
- File MCP Tool Access Disk Drive and to read and write files.
- Playwright MCP Tool Access browser on command of Claude Prompt and scrape Tally Data
- ODBC MCP Tool Connect with Tally ODBC Port 9000 and Access Tally Data
4. Tally Software: Education Mode
5. Windows Folder & Json Files : For Tally Knowledge Base

What are MCPs?
- Simple Definition: Tools that give AI models special abilities to interact with the outside world
- Technical Explanation: Extensions that allow Claude to access external systems and perform specific tasks
- Benefits: Enable AI to go beyond conversation to take real actions and access live data
- Implementation: Integrated directly within Claude Desktop for seamless workflow
Slide 7: MCPs We Leveraged:

- File MCP: Access and manage files directly from Claude
- Used to collect and organize Tally documentation
- Enabled creation of the master TDL reference document
- Playwright MCP: Web browsing and data collection capabilities
- Used for systematic scraping of Tally developer documentation
- Collected examples and reference material from authorized sources
- ODBC MCP: Direct database connection abilities
- Our "rocket engine" for TDL development
- Connects Claude directly to Tally's database structure
- Enables real-time data analysis and code generation
- Creates a natural language interface to Tally data
Slide 8: Prompt Engineering Process
From Natural Language to Expert Instructions:
- Initial requirement gathering in plain language
- Custom GPT refinement to create structured prompts
- Conversion of business needs to technical specifications
- Enhancement with TDL-specific terminology and constraints

Slide 9: Knowledge Acquisition
Comprehensive TDL Knowledge Base:
- Systematically web scraped Tally developer documentation
- Collected all publicly available TDL references
- Added reference manuals from official sources
- Incorporated real-world TDL examples from community
- Structured for optimal AI retrieval and comprehension
Slide 11: ODBC mcp Integration - The Game Changer

Slide 12: TDL Factory:-

Productivity Transformation:
- 20+ TDLs created in just 15 days
- Complex TDLs generated in single prompts
- Reduced development time from days to minutes
- Zero syntax errors in generated code
- Enhanced capabilities through AI-suggested optimizations
- Consistent implementation of best practices
Slide 13: Case Study: Turnover Analysis

Prompt In Claude Desktop:- 
Direct Analytics Results in Claude connected with Tally:



Slide 14: Why Our Approach Wins :
- Comprehensive Knowledge: Complete TDL reference embedded
- Intelligent Understanding: AI comprehends TDL structure and relationships
- Data Integration: Direct access to live Tally data
- Speed: From concept to implementation in minutes
- Quality: Error-free, optimized code following best practices
- Accessibility: Natural language interface removes technical barriers

Conclusion :
Transforming the Tally Development Experience
- Democratized Tally customization for all users
- Eliminated technical barriers to Tally extension
- Dramatically reduced development time and cost
- Enhanced quality through consistent best practices
Created new possibilities for Tally integration and
- Set new standard for Tally TDL development