AI engineer & researcher · UCLA
Sahil Kale
Building Trustworthy and Reliable AI
Hi! I am Sahil, an AI researcher and engineer working on trustworthy and reliable language models, with research spanning LLM self-knowledge, hallucination detection, uncertainty calibration, and AI-powered information retrieval. My work has been published at venues including IASEAI, SIGIR, and ACL/NAACL workshops, and I am currently pursuing an M.S. in Computer Science at UCLA following research and engineering roles at Mastercard and KnowledgeVerse AI. My main goal is to build AI systems that are powerful yet also trustworthy enough to actually help humans towards a more equitable future.
I’m open to research collaborations and exploring AI research & engineering roles for 2027. Please feel free to reach out via email to connect!
Highlights
I was invited to present my work at the WiNLP workshop at EMNLP 2026 in Budapest, Hungary.
Excited to start my M.S. in Computer Science at UCLA, focusing on NLP and safe AI.
Demonstrated my work at the U&ME LLM Unlearning workshop at ECCV 2026 in Malmo, Sweden with a full grant.
I presented as one of only ~1% of Indian authors at SIGIR 2026 in Melbourne supported by a full scholarship about low resource LLM-native web retrieval
My pitch as a start-up representative won an $800 award for KnowledgeVerse AI at ICRA 2026 in Vienna, Austria
My co-author, Kasturi presented our work on AI-backed techniques to connect farmers with policymakers at ACM DIS 2026 in Singapore.
Presented research at ACM WebSci 2026 in Braunschweig, Germany on LLM web search capabilities with an ACM grant.
My project placed 3rd after 2 rounds of heavy in-person scrutiny in the Impact Achiever category at the Zurich GenAI Awards.
I spoke at ICAART 2026 on reliable LLM systems, achieving up to 90% response stability while reducing costs by 37%.
My co-author, Dr. Luca Alfeo presented our research at ICPRAM 2026 on using knowledge graphs to improve LLM hallucination self-detection.
Spoke at the IASEAI Conference at UNESCO in Paris among Nobel laureates on how memorization can artificially inflate LLM self-knowledge.
Our geospatial modeling study was published in the prestigious Springer Discover Public Health journal.
I was promoted to Software Engineer II at Mastercard, now working on architecture serving 14M+ transactions per month.
Our paper introducing TeXpert, a benchmark for evaluating LLM-generated LaTeX was presented at ACL 2025 in Vienna
My abstract on LLM self-awareness won me a merit support award at the Pre-ACL workshop in Copenhagen.
Attended the Lisbon Machine Learning Summer School with a $1,500 Google scholarship
I presented research analysing consistency in LLM self-knowledge at the TrustNLP workshop at NAACL 2025 in Albuquerque
Our project, Avani, achieved second place in the "Pattern" track at the FAO China Youth Hackathon on Urban Agriculture
Started working with KnowledgeVerse AI as an AI Research Consultant, leading research on trustworthy and explainable LLM systems.
Joined Mastercard as a Software Engineer, working on large-scale distributed transaction and data systems.
I graduated 1st among 8,200 B.E. Computer Engineering students across the University of Pune with a 9.78/10 grade.
Our team won 1st place among 5,000+ entries at Smart India Hackathon 2023 for building a knowledge management system
Started my bachelors in Computer Engineering at the Pune Institute of Computer Technology.