AI/ML Developer | Backend Engineer | Innovator
I’m an AI/ML Developer with over 2 years at Jio Platforms Limited, specializing in generative AI, video processing, and backend systems. My journey blends creativity and technology to deliver impactful solutions.
Jul 2019 - May 2023
- Bachelor of Technology in Computer Science and Engineering.
- Focused on Data Structures, Algorithms, and Machine Learning.
Jul 2017 - May 2019
- Senior Secondary Education (HSC, CBSE).
- Achieved 87.8% with a focus on science and mathematics.
Apr 2024 - Present
- Created MCP (Model Context Protocol) servers for internal tools supporting agentic workflows.
- Developed a very lightweight intent-based search and execution system for JioCloudPC using optimized algorithms.
- Leveraged FastAPI and PyTorch for scalable performance.
Jan 2024 - Mar 2024
- Built a system to extract highlights from meeting recordings using transformer models.
- Implemented context-aware sectioning of meetings for better usability in Jiomeet+.
- Improved transcription accuracy with pre-trained VoxLingua107 models.
Nov 2023 - Dec 2023
- Designed and implemented an automated ticket assignment system for JioCRM.
- Optimized workflows using intent classification and rule-based logic.
Oct 2023 - Feb 2024
- Engineered a comprehensive 25fps lip-sync video generation platform with two novel architectures.
- Achieved a 30% increase in video quality using GAN-based lip mask generation and 2D CNN speaker identity encoders.
- Developed a multi-language speech generation system with voice cloning for 13 Indian languages using Generative AI and Stable Diffusion.
- Combined reference speaker embeddings to replicate style and tone accurately.
Aug 2023 - Sep 2023
- Contributed to Jiomeet backend by implementing REST APIs using Node.js.
- Enhanced scalability and performance for video conferencing features.
May 2022 - Jul 2022
- Revamped a Python tool for code standards verification, reducing manual review time by 85%.
- Refined the RDTF testing framework, improving stubbing for independent domain testing across 3 domains.
- Implemented 4 coding checks (e.g., banning recursive functions) and contributed to code reviews.
Apr 2024 - Present
- Built MCP (Model Context Protocol) servers for internal tools to support agentic workflows in JioCloudPC.
- Ensured high scalability and performance using FastAPI and PyTorch frameworks.
Apr 2024 - Present
- Developed a very lightweight intent-based search and execution system for JioCloudPC.
- Optimized for low latency and resource efficiency using custom algorithms.
Jan 2024 - Mar 2024
- Created a system to extract highlights from meeting recordings using transformer-based models.
- Implemented context-aware sectioning for improved meeting navigation in Jiomeet+.
- Integrated language identification with VoxLingua107 for accurate transcription.
Oct 2023 - Feb 2024
- Developed an end-to-end 25fps lip-sync video generation platform with two cutting-edge architectures.
- Improved video quality by 30% using GAN-based lip mask generation and 2D CNN speaker identity encoders.
- Built with FastAPI and PyTorch for enhanced performance and scalability.
Oct 2023 - Jan 2024
- Engineered a multi-language speech generation system supporting 13 Indian languages.
- Implemented voice cloning using Generative AI and Stable Diffusion, with reference speaker embeddings for style and tone replication.
- Achieved high-fidelity audio output with minimal latency.
Mar 2022 - Apr 2022
- Designed a real-time monitoring system for pH (6.8-7.8) and luminosity using Raspberry Pi and Arduino.
- Automated water pH adjustment and brightness control (10-20 lumen) with solenoid valves.
- Built a mobile app for 24/7 remote monitoring of water quality metrics.
Feb 2022 - Mar 2022
- Crafted an educational platform for grades 6-12 with a 10-member team.
- Deployed v2.0 with discussion forums and integrated communication tools (announcements, chat).
- Enhanced student-teacher interaction and course material accessibility.
Dive into my technical sessions and research notes: