Now shipping: full-stack AI-built engineering applications

Vikram
Pratap Singh

AI Application Engineer with a Mechanical & Aerospace core. I use AI-driven development to ship in days what traditionally takes teams months — full-stack apps, cloud deployments, and engineering tools backed by 8+ years at Honda, Suzuki & Aerotrope.

Try My Live AI App View Projects

AI Products Shipped

Years Engineering

Patents

× Faster With AI

Scroll to explore

About Me

Engineer · Innovator · Award Winner

“

AI Application Engineer with a Mechanical & Aerospace foundation. I combine 8+ years of product engineering at Honda, Suzuki and Aerotrope with AI-driven development to build and ship complete software products — solo, in weeks. I've containerized a 40-year-old Fortran solver and deployed it serverless on Google Cloud Run with an integrated LLM assistant, and built a browser-based CFD platform orchestrating OpenFOAM behind a FastAPI/Three.js stack. I don't just use AI — I use it to multiply what one engineer can deliver, end to end: concept, code, cloud, and customer.

I'm an engineer who discovered that AI changes the economics of building software. Workflows that once needed a full team — frontend, backend, DevOps, domain expert — I now deliver single-handedly by pairing deep engineering knowledge with AI-driven development. The result: working products in days, not quarters.

My proof isn't slideware. DFDC Cloud is live right now — MIT's legacy Fortran ducted-fan solver, containerized and deployed serverless with a Gemini AI copilot. CFD Studio puts a virtual wind tunnel in the browser: upload a 3D model, get meshed, solved, pressure-mapped results. Both conceived, built and shipped by one person, with AI.

What makes me different from a pure software engineer: 8+ years designing real products — Suzuki's first sunroof (team of 5), eVTOL aircraft, cryogenic LH₂ tanks, airborne wind energy — with an MSc from Cranfield University (Newell Prize, top of cohort) and 2 patents. I understand the domain the software serves.

Get In Touch Download CV

AI Application Builder

CFD Studio · DFDC Cloud — full-stack, cloud-deployed, LLM-integrated. Built solo with AI-driven development

Cloud & Full-Stack

Python/FastAPI · Node.js · Three.js · Docker · Google Cloud Run · Gemini/Claude APIs

Honda · Suzuki · Aerotrope

8+ years automotive BIW & aerospace structural design · led teams of 5 · 2 patents

Cranfield University, UK

MSc Aerospace Vehicle Design · 82%+ · Newell Prize Winner

What I Do

One engineer · full product lifecycle · AI-multiplied

AI-Driven Development

Pairing LLMs (Claude, Gemini) with engineering judgment to design, build, test & ship complete applications in days — 10× the traditional pace

Full-Stack Applications

Python/FastAPI & Node.js backends, Three.js 3D frontends, REST APIs, real-time job pipelines — from blank repo to working product

Cloud Deployment

Docker containers, serverless Google Cloud Run (scale-to-zero, ~$0/month), CI-free cloud builds, cost-governed LLM integrations

Engineering Domain Edge

8+ years of CAD, FEA & CFD (CATIA, NASTRAN, OpenFOAM) — I build software that actually understands the engineering it serves

Flagship AI Applications

Real products — designed, built & deployed solo with AI-driven development

DFDC Cloud — live ducted fan design web app

AI-Built Cloud Live Now

DFDC Cloud — Legacy Solver, Reborn

Took MIT's 40-year-old Fortran ducted-fan design code and shipped it as a modern web app: Docker container compiled in the cloud, serverless on Google Cloud Run (scales to zero, ~$0/month), Node.js API, browser UI, and an integrated Gemini AI assistant with rate & budget governance. Aerospace-grade tooling, now one URL away.

Launch Live App Case Study

AI-Built Simulation Shipped Installer

CFD Studio — Wind Tunnel in the Browser

Upload any 3D model (STL/OBJ/PLY) and get aerodynamics results in the browser: automatic meshing into a virtual wind tunnel, an OpenFOAM RANS solve orchestrated through WSL2, drag/lift coefficients, convergence plots and a pressure-coloured 3D surface rendered in Three.js. Python/FastAPI backend, shipped with a Windows installer & download page.

Product Page Case Study

Vikram
Pratap Singh