Customer profile
Customer: A US based financial services institution
Industry/Sector: Financial Services & Investment Research
Challenge: Democratizing access to complex financial analysis previously gated by a proprietary DSL requiring specialized developer expertise
Solution: AI driven, secure code generation chatbot built on modular multiagent architecture
Result: 95%+ compilation success rate, up to 98% reduction in turnaround time, and democratized access to insights

From Technical Bottlenecks to Analyst Self-Service
Financial analysis requires deep market domain knowledge, but generating production-ready visualizations in the institution’s proprietary DSL demanded developer expertise. This created a dependency loop:
- Analysts described what they needed.
- Developers spent 2–3 hours writing, testing, and validating DSL code.
- Iterations required additional cycles.
With markets moving at high speed, this workflow slowed decision-making and prevented analysts from acting on insights quickly. The institution needed a secure, scalable, AI-driven solution that would allow non-programmers to generate sophisticated, production-ready visualizations instantly.

“Rapid Circle brought the LLM expertise our team needed to accelerate development of this capability. They partnered effectively with our developers, contributing specialized knowledge that strengthened our overall approach and helped us move faster toward delivering meaningful value to our analysts and clients.” – CTO, US Based Financial Services Institution
Powered Natural Language Code Generation
AI Chart Builder enables analysts to simply describe the chart, table, or analysis they need in plain English. The system automatically interprets the request, generates DSL code, validates it against the proprietary compiler, and returns a ready-to-use visualization.
Core components:
- Azure Foundry hosting for enterprise-grade security and model deployment
- OpenAI Agents SDK enabling modular agent-based orchestration and reasoning
- Azure AI Search powering hybrid keyword + vector retrieval of DSL documentation
- Proprietary compiler integration for real-time code validation and correction
Each agent is optimized for specialized tasks, search, compilation, orchestration, making the solution scalable, flexible, and easy to extend.
Users can now ask the chatbot questions like:
“Create a table showing contribution to market gains by S&P 500 constituents over the past 10 years.”
The system fetches the right DSL components, generates compilable code, validates it, and renders the chart, within minutes.
Measurable Impact: Speed, Accuracy, Productivity
The AI powered chatbot delivered measurable improvements:
- 95%+ code compilation success rate, ensuring reliable outputs
- 98% reduction in turnaround time (from 2–4 hours to 2–4 minutes)
- Analysts self-serve advanced visualizations with near-zero developer intervention
- Instant iteration cycles (seconds vs. hours)
- Developers regain time to focus on platform enhancement rather than one-off requests
- Faster insight delivery improves market responsiveness
The organization now operates with significantly higher velocity and lower operational burden.
Ready to Supercharge Analyst Productivity with AI?
If your organization relies on specialized languages, complex internal tools, or highvolume analytical workflows, AI can radically increase speed and accuracy.
Frequently asked questions
Why was AI Chart Builder needed?
To remove the bottleneck between analyst insights and technical DSL implementation.
What made the legacy process slow?
All visualizations required manual DSL coding by developers, taking hours per request.
What improvements were achieved?
A 98% reduction in turnaround time, 95% coding success, and fully democratized analytical capabilities.
How does this benefit analysts?
Analysts focus on strategy and interpretation, not debugging or syntax. They get immediate results.
Is the solution secure?
Yes, it’s built entirely on Azure Foundry with enterprise-grade compliance and security controls.

