Portrait of Nischal Karki

I’m an NLP researcher at the Information and Language Processing Research Lab (ILPRL), Kathmandu University. I build language technology for languages that today’s AI leaves behind, especially Nepali and Tamang.

I’ve worked on three funded research projects: the Google TMT Project, a trilingual Nepali–Tamang–English translation system funded by the Google Academic Research Award; the Nepali–English Code-Mixing Project, funded by the UGC Nepal Grant; and the Election Commitment Mining Project, funded by the KU-RDI Grant.

Right now I’m adapting open source LLMs to Nepali through continued pre-training and post-training, studying data scaling in machine translation for underrepresented languages like Tamang, and exploring the safety and guardrails of multilingual LLMs.

Open to graduate opportunities, research collaborations, and engineering roles in language technology.

news

  • Two papers accepted at ICON 2026.
  • Paper on legal argument mining accepted at ICAIL 2026.
  • Co-organized the Google TMT Hackathon 2026 with 50+ teams across Nepal.
  • NepTam corpus paper at LREC 2026.
  • Recognized as Program Topper of the CS-2020 batch at Kathmandu University.

research interests

  • Low-resource NLP
  • Machine Translation
  • Code-mixing
  • Legal Argument Mining
  • Commitment Mining
  • Multilingual LLMs
  • AI Safety

selected publications

  1. LREC 2026
    NepTam: A Nepali-Tamang Parallel Corpus and Baseline Machine Translation Experiments
    Rupak Raj Ghimire, Bipesh Raj Subedi, Balaram Prasain, Prakash Poudyal, Praveen Acharya, Nischal Karki, Rupak Tiwari, Rishikesh Kumar Sharma, Jenny Poudel, Bal Krishna Bal
    Published
  2. ICAIL 2026
    Enhanced Legal Argument Mining through Parameter-Efficient Fine-tuning of Large Language Models and Retrieval Augmentation on ECHR Corpus
    Nischal Karki, Prakash Poudyal, Aayush Shrestha, Jaromir Savelka, Bal Krishna Bal, Paulo Quaresma
    Accepted First author
  3. ICON 2026
    Trilingual Machine Translation for English, Nepali, and Tamang: Corpus Development and Evaluation
    Bipesh Subedi*, Nischal Karki*, Rupak Raj Ghimire, Praveen Acharya, Prakash Poudyal, Balaram Prasain, Bal Krishna Bal
    Accepted
  4. RegICON 2025
    Benchmarking BERT-based Models for Sentence-level Topic Classification in Nepali Language
    Nischal Karki, Bipesh Subedi, Prakash Poudyal, Rupak Raj Ghimire, Bal Krishna Bal
    Published First author

in the media