2026
arXiv preprint; submitted to Interspeech 2026
Diagnostic evaluation for Indic ASR and rich transcription models for Hindi, Malayalam, and Kannada.
Arghya (Orgho) Bhattacharya
Techno Centrist
2024 -
Adalat AI
CTO and co-founder
Quit my job because I wanted to start my own company and felt like it was the right time to begin my entrepreneurial journey. Met Utkarsh, learnt about the problems of the justice system, and have been building technology to make courts more productive across the global South ever since. We are called Adalat AI. :)
2023 - 2024
Equal Identity
ML engineering lead
Led ML engineering at Equal Identity, building biometric face verification and liveness systems for mobile edge devices under tight latency and memory constraints.
2020 - 2023
Parallel work and research
work
Enterpret
Founding engineer
First hire at Enterpret, building NLP and ML systems to help our customers understand their customer feedback better. I spent a lot of time understanding supervised and unsupervised methods for creating a taxonomy from a large stream of continuous customer feedback data.
research
IIIT Hyderabad
M.S. by Research, Computational Linguistics
Research on low-resource NLP and aspect-based sentiment analysis, culminating in my M.S. thesis on advancing ABSA for low-resource languages.
2016 - 2020
IIIT Hyderabad
B.Tech (Honors), Computer Science
Studied Computer Science at IIIT Hyderabad, earning a B.Tech (Honors) before continuing into research.
bio
I am Arghya Bhattacharya (pronounced Orgho) . My formal educational training is in computer science and computational linguistics.
Back in school I identified primarily as an athlete. In college I identified as a computer science researcher. And since after college, I identify as an entrepreneur.
I use the word identified deliberately. The identities I've held have given me their value systems, and those value systems are what actually determine how I think and act.
Outside of work I take a keen interest in endurance sports, drums, philosophy, the human mind and body.
featured talks
UN AI for Good - Law Track Conference
Speaker page
Adalat AI at Stanford CodeX
CodeX group meeting, 2025
publications
Interspeech 2026 / arXiv
A benchmark and optimization study for robust Indic speech recognition in realistic conditions.
2025
Natural Legal Language Processing Workshop 2025
An India-specific benchmark for evaluating LLM legal reasoning using public legal examinations and lawyer-graded long-form answers.
Natural Legal Language Processing Workshop 2025
A voice-guided legal document navigation system that maps courtroom commands to grounded passages.
2023
M.S. thesis, IIIT Hyderabad
Thesis work on aspect-based sentiment analysis for low-resource languages, especially Hindi.
2021
4th Workshop on e-Commerce and NLP
A Hindi aspect extraction dataset and baseline study for aspect-based sentiment analysis.
2020
European Association for Machine Translation 2020
A multilingual adversarial learning approach for low-resource sentiment analysis.
CoNLL 2020
A corpus and neural baseline for one-anaphora resolution.
case studies
Stanford GSB teaching case on Utkarsh Saxena and Arghya Bhattacharya building and scaling Adalat AI's justice-technology system.
SSIR feature on Adalat AI's work addressing judicial delays with AI-powered court infrastructure.
awards and recognition
Personal recognition for work bringing AI tools into justice delivery.
Adalat AI selected for MIT Solve's 2024 Solver Class.
Adalat AI recognized for streamlining courtroom transcription.
Adalat AI named among four international technology innovations.
Adalat AI reached the global Microsoft Imagine Cup track.
India AI Innovation Challenge winner
National AI recognition from India's Ministry of Electronics and Information Technology.
Impact AI Pitchfest winner, Governance / Justice Tech
Recognition for Adalat AI's governance and justice technology work.
writing
md
media and news
2026
Regional story on Adalat AI's localized courtroom AI infrastructure, with Arghya quoted on transcription model accuracy.
Times Techies story on social-impact AI, featuring Adalat AI and Arghya's work building a justice-tech stack for Indian courts.
Profile of Adalat AI's Harvard Innovation Challenge finalist run, with Arghya named as co-founder and technical lead.
Coverage of Utkarsh Saxena and Arghya Bhattacharya being named finalists for Harvard's President's Innovation Challenge.
Media coverage of the India AI Impact Summit roundtable, quoting Arghya on people-centric AI.
India AI Impact Summit media clip.
Profile of Adalat AI and its work automating court transcription, workflows, records, and real-time updates.
Adalat AI included among organizations using AI for social impact.
2025
Ground report on Adalat AI across Indian courts, naming Arghya as co-founder and describing the 2023 founding story.
News on Adalat AI's justice-sector work beyond India, naming Utkarsh Saxena and Arghya Bhattacharya as co-founders.
Long-form interview with Arghya on building AI for public institutions and legacy court systems.
Charcha 2025 segment on Adalat AI's deployment across thousands of courtrooms.
Alumni profile on Arghya's journey from IIIT Hyderabad to building Adalat AI.
Forbes 30 Under 30 Asia profile.
News roundup of Indian Forbes 30 Under 30 Asia honorees, mentioning Arghya and Adalat AI.