Efficient Multi-Hop Question Answering over Knowledge Graphs via LLM Planning and Embedding-Guided Search
IEEE Big Data 2025 - Workshop on Knowledge Graphs and Big Data
Hello, I'm
Manil Shrestha
Building Trustworthy, Explainable Machine Intelligence
Engineer • Researcher • ML Systems Builder
Engineer and scientist powered by caffeine, curiosity, and the occasional existential crisis about super-intelligent machines.
I work at the intersection of LLMs, knowledge graphs, and ML systems, building intelligent models that think clearly—not just loudly. With 9+ years in data engineering and 5+ years in machine learning, my work focuses on making AI more trustworthy, interpretable, and grounded.
I design systems that scale, pipelines that don't break at 2 a.m., and machine intelligence that can explain itself without gaslighting you.
Research Focus
LLM Security & Trustability
Automated penetration testing, secure multiparty generative AI, and building AI systems you can actually trust.
Knowledge Graphs & XAI
Case-based reasoning, knowledge extraction, and making AI explanations actually meaningful.
Scalable ML Systems
Data backbones, distributed pipelines, and ML infrastructure that actually works in production.
Computer Vision
Synthetic image detection, vision transformers, and multimodal learning architectures.
Recent Publications
Secure Multiparty Generative AI
Workshop on Deployable AI at AAAI-2025
Towards Automated Penetration Testing: Introducing LLM Benchmark, Analysis, and Improvements
UMAP Workshops 2025 • 30 citations