Skip to main content
Kirti Jangid

Kirti Jangid

Computational Physics Researcher | AI Training Specialist & Prompt Engineer

Talent
INMember since August 20, 2026
Computational PhysicsMachine LearningMolecular Dynamics SimulationsNeural NetworksPhysics-Informed Machine LearningPhysics Tutor
·

Hire this person

Send a job offer directly to this candidate

Computational PhysicsMachine LearningMolecular Dynamics SimulationsNeural NetworksPhysics-Informed Machine Learning

About

My research spans computational physics, machine learning, and their applications to chemistry and engineering. A major focus has been the development of machine-learned potential energy surfaces using feedforward neural networks, where I designed a post-activation step that enforces low-dimensional physical constraints — ensuring, for example, that interaction potentials correctly vanish at large subsystem separations — and generated an improved set of potential energy surfaces for the 14 lowest 3A' states of O4. This work extended into advanced diabatization methods, including physically managed diabatization by deep neural network (PM-DDNN) and a dual-level parametrically managed neural network approach for learning potential energy surfaces that enable efficient molecular dynamics.

In parallel, I carried out molecular dynamics simulations of cytochrome P450 (CYP450) enzymes, characterizing them as "4D nanomachines" that sense substrate binding and drive an autonomous catalytic cycle through weak van der Waals interactions. My broader interests also include physics-informed machine learning (PIML) for infrastructure assessment and construction management — covering digital-twin updating, deterioration prediction, and multi-hazard response — as well as the integration of ethical principles (beneficence and non-maleficence) into AI-backed chemical research and development to promote sustainability.

Experience

AI Training Specialist and Prompt Engineer at Turing and Handshake AI (Present)

AI Training Specialist and Prompt Engineer with hands-on experience designing high-quality prompts and evaluating large language model (LLM) outputs across technical STEM domains. Skilled at crafting challenging, expert-level prompts that probe the reasoning limits of AI models and identify subtle errors that non-experts may overlook.

  • Designed and refined prompts to test and improve the accuracy, depth, and reasoning of large language models (LLMs).
  • Evaluated and ranked AI-generated responses for correctness, clarity, and factual reliability, providing structured feedback to improve model performance.
  • Created challenging domain-specific prompts (chemistry, physics, engineering, and quantitative reasoning) to stress-test model capabilities and expose reasoning gaps.
  • Contributed to LLM training and evaluation workflows at Turing and Handshake AI, applying prompt-engineering techniques and subject-matter expertise.
  • Applied engineering and analytical background to build rigorous, technically demanding prompts that support the development of safer and more reliable AI systems.

Education

Doctoral Studies (Ph.D.) in Computational Physics – The University of Texas at Dallas (2022-08 – 2026-08)

BS–MS in Physics – Indian Institute of Technology (IIT) Mandi (2014-07 – 2020-08)

Skills

Computational Physics ResearcherAI Training Specialist & Prompt Engineer

Reviews

Related