Skip to content
View Danny-r-2000's full-sized avatar

Block or report Danny-r-2000

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Danny-r-2000/README.md

Danny Rajalingam

Oxford DPhil in hypersonics, now building Ember — AI to augment the physical and human-labor workforce.

My thesis is that the next wave of useful AI for the real world is inverted: the assistant speaks first, without being asked, based on the state of the work in front of the person. The old model — type a question, read a wall of text — does not survive contact with a bricklayer whose hands are covered in mortar or a turbine engineer trying to keep a string line straight. The two pinned repos below are my working proofs of that thesis, each tied to one half of my background.

Featured projects

hypersonic_boundary_layers  —  a physics-informed neural network toolkit for hypersonic boundary-layer heating, wrapped in a proactive research copilot. A semi-supervised PINN for the Blasius equation with a custom output transform beats a data-only MLP by 4.4× RMSE on the shear profile using only eight sparse samples. Pure-NumPy autograd, no torch dependency, installs in under five seconds. Python · PINNs · Hypersonics · CFD

ember-rocky  —  the reference implementation of Ember's inverted-UX thesis. A ~90-line event-driven proactive loop with pluggable role triggers that produce different guidance for different people from the same event stream: a novice bricklayer gets a step-by-step brief, the foreman gets a one-sentence crew rollup, the safety officer gets an audit-ready incident line. Runs end-to-end with zero credentials and zero hardware. Python · AI-agents · Voice-first · Construction-tech

About me

I did my doctorate on hypersonic boundary-layer transition and the physics of shock/boundary-layer interaction, which is mostly a story about how to get useful numbers out of a regime where your measurements lie to you and your simulations disagree with themselves. That background is why I think Ember's bet — AI copilots for environments where the user cannot stop, type, or read — is the right shape of problem.

  • 📫 Reach me at ramondannyrajalingam@gmail.com
  • 🧭 LinkedIn
  • 🔥 Ember is pre-seed. If you work on construction, warehouse, field-service, or enterprise onboarding problems and want to compare notes, say hi.

Pinned Loading

  1. hypersonic_boundary_layers hypersonic_boundary_layers Public

    physics-informed ML for hypersonic boundary layers. Pure-NumPy autograd, semi-supervised PINN for the Blasius equation, Van Driest II skin-friction module, airfoil surrogate, proactive agent wrappe…

    Python 1