Streamlit announced that Streamlit 1.0 is generally
available. The open source project has more than 16,000 GitHub stars, has been
downloaded more than 4.5 million times and is used by more than 10,000
organizations, including more than half of the Fortune 50.
Streamlit is a powerful and easy-to-use framework that lets data scientists
quickly build web apps to access and explore machine learning models, advanced
algorithms and complex data types. These apps are used for everything from
advanced analytics dashboards to sales and marketing tools based off of the
latest predictive algorithms. Streamlit's unique workflow is 10x faster
than other alternatives, making it possible for data scientists to go from idea
to deployed app in only a few hours.
Compared to existing dashboarding tools, Streamlit offers a far greater range
of AI-powered use-cases: - Apps that combine different
APIs to produce new insights for sales, marketing and product
- Apps that surface custom
product recommendations for sales people to pitch
- Data exploration apps that
let non-technical users explore SQL data via AWS S3, BigQuery, Snowflake or any
internal or cloud database
- Data labeling apps to
annotate new types of data for machine learning or data cleaning
- Apps that take data on user
cohorts and generate targeted promotional coupons
- Fraud detection apps to
understand how different detection models trade off detection rates and total
dollar loss
- ROI tools that help
companies land deals by showing what if analysis to customers
- Apps for procurement teams
to know what types of inventory to buy based on the latest trends
- Apps for optimizing
operations such as where to place electric bike chargers for the highest
utilization
- Model comparison apps for
ML teams to compare models side by side to see performance changes
"Streamlit
shows the incredible power of community-driven software to disrupt industries,"
said Adrien Treuille, co-founder and CEO of Streamlit. "Declaring 1.0 is a
major achievement for Streamilt's incredible development team and
community. And we're just getting started. What we've done in the past two
years is nothing compared to what's coming next!"
Streamlit users love Streamlit for the flexibility and power that it provides: - Joanna Tang, Senior Data
Scientist at Squarespace, said: "Streamlit is great because it lets
non-technical audiences understand the metrics we develop."
- Tyler Richards, Data
Scientist from Facebook Research who recently authored a book about Streamlit,
said: "Streamlit makes my work so much more impactful by helping me quickly
turn my analyses into beautiful, dynamic web apps."
- Charlie Lefrak, Software
Engineer from Mapbox, said: "I use Streamlit for everything from spinning
up prototypes, exploring and interrogating data and creating finished
business analytics products."
- Chinmay Gaikwad, Senior
Systems Engineer at Infosys, said: "Building and deploying an ML model
have never been easier. Thanks to Streamlit, it makes building web apps
super easy and I would highly recommend it for data science projects."
- Amit Marathe, Director of
AI & ML at Inseego Corp, said: "If you want to impress your boss with
a quick data visualization dashboard, you can build a simple and beautiful
data visualization dashboard in a couple hours using Streamlit!"
- Eric Sims, Sr. Analyst,
Strategy & Analytics at LendingTree, said: "Deploying on Streamlit was
ridiculously easy - took 5 minutes tops, probably less, actually."
Streamlit 1.0 has added many powerful features since it was
first introduced:
- Improvements to app speed
and responsiveness: Improvements to caching, switching to Apache Arrow for
serialization and memory management updates have resulted in major gains
for speed and responsiveness
- Customization with app
layout primitives and theming: Columns, sidebar, wide mode, expanders,
dark mode and customizable themes mean that users can now lay out their
apps and match it to their ideal style of company brand
- Ability to create complex
apps by adding statefulness: With the addition of Session State and
forms, users can now choose when their app re-runs and it lets users do
complex things like pagination, annotation or even creating games
- An amazing ecosystem of
components and integrations: Users can extend their apps even further by
writing their own components or use ones from the community
including integrations, more libraries like SpaCy, HiPlot, Folium and
Observable and new functionality like sending and receiving video or drawing on
a canvas
Streamlit's 2022
Roadmap
Today Streamlit is also sharing its roadmap for 2022 for Streamlit: https://share.streamlit.io/streamlit/roadmap.
- Make Magical
Apps: Streamlit has already made it 10x faster to make great apps but now
it wants to make those apps even better. That means an unbeatable set of
widgets -- everything from sortable/filterable/editable databases and
tables, clickable charts, image selectors and editors, amazing audio and video
players and uploaders, and a lot more options for layout and customization
- First-Class Developer
Experience: Streamlit wants everything about coding a Streamlit app to be
an amazing experience. So it will be working on everything from making it
easier to connect to data sources, easier to cache and interact with data,
and easier debugging
- Enhance the Viewer
Experience: Ultimately users are developing apps for someone else to view
them so it also wants to help them make great apps for their viewers. It is
thinking about how to give a viewer experience distinct from the developer
experience, so viewers can more easily understand and interact with the app and
give direct feedback
- Rapidly Expand the
Ecosystem: The community has already given so much to Streamlit and it
wants to make it even easier for users to share code, components, apps,
and answers -- so it is going to be launching new features that make it
easier to get started with new apps, find code snippets, search for the right
component to add to apps, engage with the worldwide community, and get
recognized for what users contribute to the community
|
|