Low-Code Scalable, Hosted D2i Apps: Data-Driven Interactive Dashboards

Build Product Prototypes in a Low-Code Environment without a Team

by Arth Dharaskar

The Pain. Setting up a scalable, hosted Data-driven web app is a tedious, time consuming process requiring contributions from a team of backend/infrastructure engineers, software/systems engineers and front-end engineers and product designers alongside data engineers and machine learning engineers/data scientists.

As a result, developing data-driven products is resource intensive, time consuming and process-ridden within any organization. Subsequently, experimentation with product design is extremely limited. Therefore, the time to market for bringing innovative product design is either extremely long or never happens.

What is needed? A low-code platform for a product person to quickly iterate with product ideas by experimenting with them in a production environment. There are solutions that tackle aspects of this problem: cloud service providers (AWS, GCP, Azure, etc.) enable access to commodity compute and data storage for backend and infrastructure engineers; platforms for machine learning and data science (TensorFlow, Databricks, etc.) provide access to machine learning infrastructure for data scientists; programmable dashboards (Dash, for example) enable building data-intensive interactive dashboards for front-end engineers; and wire-framing platforms (Figma, etc.) enable building designs for product designers. However, none of these enable a product person, on her own and without a team, to get all of these together to build a production prototype — this is urgently needed.

The Solution. With Ikigai, one can build data-driven interactive web applications in a low-code environment. In effect, using a variety of connectors, data from first party or third party sources (including web scrape and all that) can be accessed using a low-code computation environment. It is feasible to prepare, process, analyze, predict and optimize using data in a secure environment, using a low-code dashboarding environment. It is also possible to bring a wire-framed design to life that is intimately connected with computational tasks to enable interactivity. Finally, such dashboards can be published as a hosted web app to bring a web application to life; and indeed, it can be easily integrated into an existing web app stack through appropriate URL redirection to enable experimentation.

To explain this, let us take an example of building a movie recommendation system using live interaction data. The process will be divided into three steps. One, connecting to data. Two, processing data, adding recommendation algorithm and enabling interaction. Three, building a dashboard that interacts with the earlier steps to both enrich data, compute recommendations and provide an interactive experience to the end user. Finally, we can publish the dashboard as a web app.

Step 1. Connecting to a data source in Ikigai is as simple as dragging and dropping our Imported Facet! In this case, we point it to the MovieLens dataset.

Step 2. Here we show the final pipeline after connecting the remaining facets. It consists of using our Recommend Facet to produce collaborative-filtering based recommendations and the remaining facets provide metadata [Movie Title, Similarity] for our predictions to enhance output interpretability.

Step 3. Finally, we create the D2i-App powered by the flow shown in the prior step that allows users to have an interactive experience with the data!

This final app can be shared easily with stakeholders within/outside the organization with the click of a button, using anything else would be — as Spock would put it “Highly Illogical”.

About the Author

Arth Dharaskar (Applied Data Scientist)

Arth Dharaskar is an Applied Data Scientist at Ikigai Labs where he is working on building a scalable machine learning infrastructure. He holds a Master’s degree from the University of California — San Diego specializing in Machine Learning and Data Science and a Bachelor’s degree in Computer Science from Rutgers University. When he is not working, you will find him lifting weights or reading books!




AI-charged spreadsheets for data operators who run mission-critical biz processes using data.

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Manage Azure Event Hubs with Azure Service Operator on Kubernetes

Transfer Large Files Using a LicenseSearch APIs

It is possible to act quickly in your e-commerce by using the weight calculation module

Pauper Set Review: Innistrad Crimson Vow

TIL You can deploy WebGL Unity games on OpenSea as a NFT

Elasticsearch for beginners

10 Python Built-in Functions Every Data Scientist Should Know

The Internship Experience At Let’s Grow More.(LGM)

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Ikigai Labs

Ikigai Labs

AI-charged spreadsheets for data operators who run mission-critical biz processes using data.

More from Medium

How to sync data between two CRMs

Choosing an Analytics Tool. Metabase Vs Superset Vs Redash

Gilbane Advisor 5–4–22 — web renaissance, MDS ecosystem

An early-stage founder’s guide to building out your data stack