At SiddhVesh Consultancy, I'm building the AI Studio workflow engine within Hermes, an enterprise multi-tenant Data & AI platform. The work spans a React canvas UI, Strapi schema design, and Python backend services.
I architected a single Start-node trigger engine for manual, webhook, scheduled, and agent-tool workflows. I also implemented multi-parent DAG join logic that contributes to a distributed executor supporting 150+ plug-and-play components.
I integrated Google Workspace connectors as configurable workflow nodes and delivered the feature across hermes-frontend, hermes-cms, and ai-studio-service while maintaining backward compatibility with legacy studio types. I also used Socket.io to communicate live execution events between the frontend and backend.
Earlier, I developed a Groq API chatbot at MSBC Group, NLP sentiment and churn prediction models at Maxgen Technologies, and a Flask web application and image-classification system at Netwise Consultancy Services. My work has included RAG LLM projects, with reported model accuracies of 77%, 87%, 95%, and over 90% across these projects.
