COVID-19 Testing at scale

•

Simple & Pretty — Product design & strategy for startups

AI software platform to accurately and rapidly detect viruses

Background

Shortly after the COVID-19 outbreak, Deep Innovations started to build and train a vision-based, deep neural network algorithm that could identify viruses by using cloud-connected SEM microscopes in a spit sample.

Objective

Envision a scalable, flexible, TAAS (Testing As A Service), which would facilitate rapid, low-cost virus testing. For use at health centres, transport stations, public venues, anywhere where a negative result is required for access for public safety.

Solution

In order to deliver a solution to market faster, we divided it into 2 phases:

Phase 1: Cloud-based, software solution, for use medical centers, hospitals, testing labs. Leveraging existing processes and hardware

Phase 2: A brandable, self-service kiosk for automated, on-location virus testing

Deliverables

POC, Customer journey mapping, Software & Hardware concepts

Project overview

We began by researching common medical testing procedures, technologies and hardware. This included speaking to medical professionals and lab assistants.

This allowed us to map out the end-to-end service lifecycle, including both the physical and digital interactions amongst the different actors (testee, test assistant, lab assistant, SEM operator, health consultants, organisation administrators etc)

Working closely with our team, we wrote the MVP Product spec and Strategy documentation to clearly define the product solution, before conceptualising the UI.

We worked with the developers to define a proof of concept (POC) to rapidly build and explore the core functionality. We defined 5 levels to the POC, starting with recreating our internal tech demo in app, then each level introduces another piece of functionality e.g. capture image on SEM microscope, multi-SEM processing, multi-user, text scanning and tracking.

Concepts

The system was composed of 4 main areas, which area a user can access depends on their user type and access level:

"Having worked with Matt and Judit for 3+ years, they are a great team to work with, very productive, and contribute a lot to the quality of the product, delivering best results from proposal to production. Lots of ideas!"

Faris Dabour, Co-founder & CTO at Deep Innovations