Shailesh Mishra
Scene Reconstruction Engineer at Quidient
Kathmandu, Bagmati Pradesh, Nepal
Summary
🤩
Rockstar🎓
Top SchoolShailesh Mishra is a Scene Reconstruction Engineer with eight years of experience building 3D reconstruction, animation, and applied AI systems across research and industry. He has driven core algorithm development at Quidient and delivered research-grade results at DFKI and NVIDIA, including a Eurographics paper and substantial speedups in distributed training. His background spans sign language animation, inverse kinematics, diffusion models, LLM-enabled RAG systems, and real-world security research—he even contributed examples of LLM attack vectors in a notable open-source repo. Comfortable moving between Blender, Three.js, and production ML stacks, he combines deep technical rigor with practical engineering that accelerates pipelines and improves animation fidelity. Based in Kathmandu, he brings an uncommon mix of generative-AI research, security-minded thinking, and hands-on graphics expertise.
8 years of coding experience
3 years of employment as a software developer
Bachelor's degree, Computer Engineering, Bachelor's degree, Computer Engineering at Tribhuvan University, IOE, Pulchowk CampusMaster of Science - MS, Computer Science, Master of Science - MS, Computer Science at Universität des Saarlandes
Github Skills (8)
prompt-engineering10
security-testing10
vulnerabilities10
exploit10
python10
vulnerability10
chat-application9
chat9
Programming languages (7)
Github contributions (5)
greshake/llm-security
Feb 2023 - Feb 2023
New ways of breaking app-integrated LLMs
Role in this project:
Security Engineer
Security EngineerContributions:18 commits, 2 PRs, 1 push in 2 days
Contributions summary:Shailesh primarily focused on demonstrating and expanding attack vectors against LLMs, specifically targeting a chat application. Their commits introduced examples of data exfiltration techniques, including crafting prompts to extract user information. They also implemented a "spreading" mechanism, showcasing how an attacker could propagate malicious instructions to other users via email. The user's work involved modifying application logic and integrating tools for demonstrating security vulnerabilities.
shhra/anim
Oct 2021 - Mar 2022
Contributions:66 commits, 2 pushes, 3 branches in 5 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.