Open To Research Collaborations

NLP research, audio intelligence, and systems built with signal and precision.

I am Indrayudh, an NLP researcher at Lingo Labs, IIT Gandhinagar and, since January 2026, also associated with Deccan AI, where I work on practical evaluation systems for modern AI workflows.

View Projects Contact Me LinkedIn
Indrayudh's Profile Picture

Current Focus

ASR, Agent Evaluation, VLM Benchmarks

M.Tech in Artificial Intelligence (2024-2026)

Role

NLP Researcher, ML Intern

Affiliations

Lingo Labs, Deccan AI

Base

IIT Gandhinagar

About Me

I am Indrayudh, an NLP researcher with a strong passion for Artificial Intelligence. At IIT Gandhinagar, I work under the guidance of Prof. Mayank Singh on research that sits at the intersection of speech, language, and deployable AI systems.

My work spans audio models, Automatic Speech Recognition for Indic languages, and evaluation research for language and vision-language systems, with an emphasis on building robust pipelines that turn research ideas into usable tools.

Research Themes

Automatic Speech Recognition Audio Models Cross-Lingual Semantics Indic Languages

Education

M.Tech in Artificial Intelligence

Indian Institute of Technology Gandhinagar

2024 - Present

Research Focus: NLP, ASR

Advisor: Prof. Mayank Singh

TA: NLP (Fall 2024)

TA: PSDV (Spring 2025)

Masters in Computer Applications

Chandigarh University (CU-IDOL)

2022 - 2024

Bachelors in Computer Applications

Maulana Abul Kalam Azad University of Technology (WBUT)

2019 - 2022

Experience

ML Intern

Parallel Role

Deccan AI

Jan 2026 - Present

  • Worked on trajectory-based workflow evaluation using a rule-based plus LLM-as-a-judge approach.
  • Developed sanity and contamination checkers for Terminal-Bench tasks to improve evaluation reliability.
  • Worked on a spatial counterfactual occlusion rearrangement evaluation benchmark for VLMs.

NLP Researcher

Lingo Labs, IIT Gandhinagar

2024 - Present

  • Building an ASR system for Indic languages from scratch.
  • Developed a cross-lingual STS model that processes audio from an Asian tonal language to English through a multi-stage pipeline.

Software Developer

Wipro Technologies, Bengaluru

2022 - 2023

  • Transformed legacy COBOL modules into efficient JCL scripts for RSA Insurance Group systems, reducing processing time by about 40%.
  • Reduced defect leakage by around 30%, improving compliance and software quality.

Projects

Selected work in research, tooling, and applied reasoning systems.

In Progress ASR

ASR for Indic Languages

Python, PyTorch, Transformers

A speech recognition system focused on Indic languages, designed to push toward robust multilingual audio understanding.

Learn More →
Reasoning LLM

CoT-Llama3.3-70B

Python, PyTorch, Prompt Engineering

Implemented a Chain of Thought reasoning framework using Meta's Llama 3.3 70B model through the Groq API.

View Repository →
Translation Speech

Cross-Lingual STS Model

Python, Transformers, Vosk, PyTTSx3, DeepFilterNet, Tacotron2-DDC

Built a high-performance speech-to-speech translation framework that handles tonal language audio and produces English output through a multi-stage pipeline.

View Project Details →

Skills

Natural Language Processing Automatic Speech Recognition Machine Learning Deep Learning Python PyTorch TensorFlow Transformers Data Preprocessing Model Evaluation Research and Development Statistical Analysis Git and GitHub

Get In Touch

Research collaborations, interesting ideas, and AI conversations are always welcome.

I'm always open to discussing research opportunities, collaborative work, or connecting with people building thoughtful AI systems.

Email Me Connect on LinkedIn GitHub