Prashanthan Ganeswaran Projects & ai / making Speaking experience

Jump to the speaking record

First Touch

Instant, easy prioritised to-do list

  • Applied AI
Never forget an important idea / task

Problem

Losing track of important tasks & ideas that come to you when you can’t quickly capture them in a to-do / notes app.

Solution

A system to quickly capture tasks & ideas using a [nearly], always accessible device - smart watch, by recording them by voice whilst in the moment and forgetting about them. They are immediately processed in the background and then prioritised into a to-do list accessible on the custom watch, phone and desktop apps, as well as Notion.

Come Fly With Me

Flight info on smart glasses

  • AR / VR
Utilising AR/smart glasses to keep important flight information easily accessible

Problem

Whilst navigating the airport, I want to make sure I’m up to date with flight changes but don’t want to be glued to my phone the whole time. It also allows my hands to be freer to carry luggage etc.

Solution

Utilise the “heads-up-display” nature of the Meta Ray-Ban Display glasses to allow me to quickly load my flight details, see the gate number and even load connecting flights, all in one place. This allows my phone to stay in my pocket and my hands to be freer.

Operation: Life Admin

Paper records, digitised and searchable

  • Applied AI
Capture, convert, sort, search & analyse

Problem

Years of paper records, letters, car documents and bills sitting in drawers. None of it easily accessible, searchable or ready to be analysed to be optimised.

Solution

All records, letters sent in the mail, official documents for my car, bills etc. digitised by a local LLM, converted into .md files with the original layout and automatically categorised and stored. Suddenly years of data are quickly accessible to AI models for quick access and complex analysis.

prashhh

Local dictation, faster than typing

  • Applied AI
Rapid transcription. Complete privacy.

Problem

Typing slows down transfer of data from a person to their machine. Transcription services exist but legacy versions were not very accurate and cloud-based AI models require internet access and a subscription.

Solution

Dictation using OpenAI’s Whisper model locally to quickly transcribe speech to text when working on my PC for a faster workflow than typing, especially when prompting. No lag or security concerns, whilst also working offline.

AR tech demonstrator

Variable focus AR DISPLAY DEMO

  • AR / VR

Problem

When using AR / smart glasses with waveguide display technology the virtual content that you see projected in front of you, is focused in the far distance. So when you look a your hand, you cannot see both your hand and the content in focus at the same time. This can strain your eyes and doesn’t feel natural.

Solution

Utilising Oxford Optical Labs fluid lens technology, we’re able to change the focal plane of the display dynamically. I created a demo experience that showcases first the problem with today’s single focal plane displays and then shows off the variable focal plane display tech we’ve been developing. During development I used AI to help with the demo routine code, that drives virtual content and lens control, as well as automating Blender batch renders to provide the virtual content.

Car media controller

The half the car was missing

  • Applied AI

Problem

The car has an aux socket and no Bluetooth. Skipping a track means picking up the phone and looking away from the road to do it. This is not legal in the UK!

Solution

Using a small ESP32 module and a simple 5-way joystick, I built a physical control allowing you to change track, play and pause, done by feel without looking away from the road. It turns on when the car is started, immediately connects to my phone and the controls work with no perceivable lag. Sometimes physical buttons are better than screens.

There’s more…

  1. 3D printed animations
    Art
    A novel approach to stop-motion animation whilst 3D printing.
    See the work on YouTube ↗
    16M+ views
  2. prashanthan.co.uk
    Applied AI
    Design sketched out, digitised using Affinity, with AI writing the code and scripts measuring the result back against the original design.
    2026
  3. Amoxy
    Applied AI
    An app that plans and gamifies taking antibiotics on time, to keep a consistent dose throughout the course.
    2026

Speaking experience

  1. AWE USA
    Long Beach
    Prescription on Demand

    Frictionless correction of optical prescriptions is a key enabler for the adoption and market growth of XR. It is essential for providing a clear, comfortable XR experience to over half of the market, demonstrating at point-of-sale, enabling home sharing and enterprise applications, and maintaining resale value in the headset. Our solution provides a perfectly corrected experience within seconds using a fluid-lens technology that delivers full spherical and astigmatic correction in a compact, integrable module. The same technology also creates more immersive experiences by enabling variable depth of the focal plane.

  2. Printed
    Amsterdam
    Makers Central
    Birmingham
    3D Printed Animations Uncovered

    The story of how and why prash decided to connect a mini computer, lights and cameras up to his 3D printer and photograph his prints, mid-print, to create a new medium of stop-motion animation. Syncing this to audio, his creations involved prints from Stranger Things, Kendrick Lamar songs or sometimes just replies to comments on videos!

    33 min
    Printed, Amsterdam
    45 min
    Makers Central, Birmingham
  3. SPIE AR/VR/MR
    San Francisco
    Breaking the near-vision barrier with focus tunable lenses

    Focal rivalry presents challenges for the many new applications utilising hand tracking to stimulate interaction with virtual objects. Focus tuneable lenses can be used to move the focal plane of the display to match the plane of accommodation.

  4. SXSW Interactive
    Austin
    Selected as speaker for five consecutive years

    SXSW Interactive festival speaker, across meet-ups and a co-presentation with HTC Vive. Started off in the field of Brain-Computer interfaces and then moved over to AR/VR in 2018.

  5. NEXT17
    Hamburg
    Brain-computer interfaces and what comes next

    Mapping out the current state of Brain-Computer interfaces in 2017, highlighting the current major players in the market and looking to the future, predicting what possible outcomes there are for the technology in the near and far future.

Get in touch