About project
Symphony is a communication and markets technology platform for financial institutions and other regulated organizations. It brings secure collaboration, market workflows, and financial intelligence into a connected digital environment.
The product is designed for professionals who need to exchange sensitive information, access market insights, and coordinate work across organizations while meeting strict operational and compliance requirements.
By combining trusted connectivity with analytics and information services, Symphony helps users work with financial data and conversations in one ecosystem and reduces fragmentation between communication and research tools.
Duration
08.2023 – 07.2024
Technologies
Here’s the Result
- users received faster and more responsive analytics and administrative interfaces.
- meeting conversations can be transcribed and summarized into questions, decisions, tasks, and other key points.
- conversation analysis provides additional context through emotional tone, keywords, and key phrases.
- teams can review important meeting information more quickly and focus on relevant actions.
Here’s What We Did
- optimized the platform with server-side rendering, static generation, and lazy loading.
- developed the conversation capture and transcription pipeline with Amazon Transcribe.
- integrated GPT-4 to extract questions, decisions, tasks, and key points from transcripts.
- implemented emotional-context analysis and keyword extraction with Whisper and RAKE.
- added automated testing and CI/CD workflows for reliable releases.
Roadmap
Performance Optimization and UI Enhancements
We optimized rendering and loading behavior, improved analytics and administration interfaces, and added automated tests. This stage was important for establishing a responsive and reliable foundation before introducing real-time processing.
Real-Time Conversation Capture and Transcription
We built the conversation capture and transcription pipeline and added GPT-4 processing for questions, decisions, tasks, and other key points. This stage turned recorded discussions into structured information users could review immediately.
Emotion Detection and Contextual Analytics
We introduced emotional-context analysis and real-time keyword extraction with Whisper and RAKE. This stage added more context to transcripts and made important topics easier to identify.
We completed end-to-end testing, scalability and performance optimization, front-end integration, and deployment preparation. This stage was important for delivering a stable production workflow and supporting user onboarding.
Main Functionality
Integrated Financial Intelligence
Global Financial Connectivity
Company Culture and Careers
Leadership Team
Contact Us
