Manglesh Patidar

• Developed backend services and REST APIs for the MyoConsult platform using C# (.NET), Entity Framework Core,
SQL Server, and Azure Blob Storage, supporting scalable healthcare data management, consultation workflows, and
report generation.
• Developed an identity-preserving NVS pipeline using LAM to generate front-view faces from dual side-profile images,
with InsightFace InSwapper for enhanced facial consistency.

• Developed an end-to-end AI-powered soccer highlight generation system using FFmpeg, OpenAI Whisper, Gemini
LLM, and Streamlit, automating event detection, timestamp extraction, and highlight reel generation from full-length
match videos.
• Designed and implemented a multimodal AI pipeline integrating speech-to-text, audio processing, LLM-based event
extraction, and automated video clipping, enabling scalable sports analytics and highlight generation.

• Developed a production-ready e-bike infotainment HMI dashboard using LVGL, C/C++, and Microsoft Visual Studio,
delivering a modern embedded user interface as an open-source alternative to commercial HMI platforms.
• Designed and implemented interactive embedded GUI components using LVGL Editor, reducing licensing costs by
Rs.10 lakh/year while eliminating source code disclosure and vendor dependency constraints.

• Designed and developed Android Automotive applications using Java, AOSP, MVVM, and Figma, optimizing UI/UX
workflows and reducing user interaction time by 15%.
• Implemented Bluetooth PBAP and integrated external APIs within AOSP, enabling seamless synchronization of calls,
contacts, and favorites while supporting real-time weather information and feature validation through Android emulators.