Hi I'm Prithvi Prabhu.
The purpose of computing is insight, not numbers
- Richard Hamming
Delivering insights to those that need them, in a usable form, is still cumbersome in 2022. Most analytics / ML / data science projects never make it to production.
Quick prototyping, continuous feedback from end-users, and rapid improvement is of the utmost necessity. Resorting to long-winded product design/development processes and client/server architectures is time-consuming and introduces accidental complexity, mostly avoidable.
There are simpler ways. This blog explores how.
About #
I work at the intersection of computing and design to explore and invent new ways of interacting with information. I author software that makes it simple, fast, fun and easy to discover, interpret and communicate patterns in data.
I am interested in:
- Visual programming, no-code / low-code / self-service / rapid application development systems.
- High-performance graphics, real-time information visualization, data visualization, visual analytics and self-service decision support systems.
- Research and development of novel HCI techniques for interactive exploration and analysis of large-scale data.
- Patterns, practices and programming strategies for building robust, large-scale, highly interactive analytical systems.
- Symbolic computing, interactive computing, distributed systems, compilers and DSLs.
Work #
I've founded or been an early engineer in three startups.
I currently work at H2O.ai at Mountain View as Chief Technologist, a fancy title for "mostly programmer, programming" (12,000+ commits in Github since March 2014).
I lead the development of H2O Nitro and H2O Wave, open-source rapid application development systems for authoring interactive real-time information displays and analytical decision-support systems.
I consider myself incredibly fortunate to be working with some of the smartest people in the industry, including Leland Wilkinson of Grammar of Graphics fame, and hacker extraordinaire Cliff Click.
See Projects for a complete list of things I've worked on.
Programming Languages #
I like tinkering with programming languages.
- Sponsoring: Zig.
- Actively using: Go, Python, Typescript, Dart.
- Used in the past: ActionScript (Flash, Flex), C, C# (WinForms, WPF, Silverlight, ASP.NET), C++ (VC++, MFC), CoffeeScript, F#, Java (J2EE, Swing, AWT), Javascript / Node.js, Lisp, OCaml, PHP, Perl, Ruby, Rust, Visual Basic.
- Planning to learn: J, Nim, Erlang/Elixir and Red.
- Fascinated by: Postscript, Forth, Joy, K and Q.
- Hacking on: Zero - a browser-based concatenative interactive graphics language that I hope to release some day.
Patents #
- Evolved Machine Learning Models - (Pending) Feb 27, 2019
- Systems and methods for interest-driven data visualization systems utilizing visualization image data and trellised visualizations - Apr 3, 2018
- Systems and methods for interest-driven data visualization systems utilized in interest-driven business intelligence systems - Nov 21, 2017
Music #
When I'm not working, I'm usually into sample-based music production, mainly using an MPC Live II, an Octatrack mkII, a Roland FA workstation keyboard, and a couple of Novation Launchpads hooked up to Ableton Live. The MPC Live, in particular, is a brilliantly designed sampler and endless fun. I've spent countless hours studying and reverse-engineering sampling and sequencing techniques used in boom-bap, hip hop, trip hop, lo-fi and related genres. Someday I hope to design and build an inexpensive, self-contained, compact portable hardware sampler like the MPC, embedded systems and all. Email me if you want to collaborate ;-)