Why Mage exists

May 19, 2021 · 9 minute read

Tommy Dang

Engineering

Xiaoyou Wang

Engineering

TLDR

Mage makes it easy for

product developers

to use AI and their data to make predictions. Use cases include churn prevention, product recommendations, customer lifetime value, ranking, and more.

Mage is designed and built for

product developers

with no prior AI experience. Developers can get started with AI quickly and continue learning with Mage. As they become more proficient with AI, our platform unlocks greater customization and flexibility for the individual.

In ancient times, a 

mage

(pronounced like page but with an M) was someone who used magic to achieve an objective or acquire knowledge. The modern world has produced a different kind of magic: advanced technology in the form of artificial intelligence (also referred to as AI).

However, harnessing this power can be difficult and reserved for a select few. Here at 

Mage

, we equip companies with accessible AI technology so they can transform the lives of their customers.

Our mission

We believe in a world where every business can create magic in the lives of their users. That’s why

our mission is to equip businesses with accessible AI technology so they can deliver transformational customer experiences

.

While working with hundreds of engineers across 30 plus teams at

Airbnb

, we saw that product developers had innovative ideas on using AI to improve their user’s experience. However, these ideas rarely made it into product because developers depended on data science resources, which weren’t readily available to them, to help implement those ideas.

Harness the power of AI

There are lots of great tools available for building AI models. Large companies, like Airbnb, even build their own AI tools internally. However, those tools were designed and built for people who are already experts in AI: data scientists and machine learning engineers.

Product developers know what AI is and how it can be used. Now, they just need a tool that’s designed for them and tailored towards their development process and level of familiarity with AI. We created Mage to solve this. Mage makes it easy for product developers to use AI and their data to make predictions. Use cases include churn prevention, product recommendations, customer lifetime value, ranking, and more.

Who we serve

Mage is designed and built for product developers with no prior AI experience. Developers can get started with AI quickly and continue learning with Mage. As they become more proficient with AI, our platform unlocks greater customization and flexibility for the individual.

Mage is a collaborative tool where multiple people across multiple teams can contribute data, refine data through custom annotations and transformers, fine tune AI models, etc. The more people that use Mage together, the better the model’s predictions will be.

We are focused on serving the needs of companies with 250 employees or less. Companies of this size typically have little to no AI expertise, no existing solutions tailored for their company size, and these companies are the underserved market.

Current AI tools are designed for and sold to large enterprises because companies of that size typically have more money, more AI resources, and a perceived greater need for AI. However, this makes AI tools inaccessible to smaller companies whose need for AI is just as great and who may not have as many resources as a large company. This is why Mage provides an enterprise-grade AI tool for small to medium sized companies.

How we do it

Mage works in 3 steps: connect your data, train your model, make predictions.

1. Connect your data

There are 3 ways to connect your data: uploading a CSV file, streaming data via an API, or connecting Mage directly to your database, data warehouse, or data lake. We provide tools for annotating your data and transforming your data. Behind the scenes, Mage will automatically take care of some data cleaning and common aggregations.

2. Train your model

Mage makes it easy to create models with very little code. First, you choose the use case for your model (e.g. ranking, predicting churn, categorizing, etc.), then you select which data your model learns from. Once you’re done with that, Mage takes care of the rest: cleaning the data, building the training set, adding more features (aka feature engineering), selecting which feature to use, selecting an algorithm, tuning

hyperparameters

, training the model, and deploying the model.

Once the model is trained, Mage will explain how the model was created, what features are most important, the metrics for how the model performed during training, and suggestions for improving the model.

There are 3 ways to use your model once you’re ready: download all the predictions to a CSV file, automatically export the predictions to your data warehouse on a one-time or recurring basis, or make an API call to Mage from your backend server, front-end app, or native app for real-time predictions.

What we believe in

We are a team of dedicated and customer-obsessed engineers, designers, and marketers with experience ranging from early stage startups to large tech companies like Airbnb. Our team has spent most of our careers building tools that empower people to do things they’ve never been able to do before.

Values are aspirational. Virtues are values in action. Here are the 6 core virtues that define who we are and how we behave.

Victorious as a team

One for all and all for one (source: Marvel Studios)

Everyone at Mage is selfless and puts the needs of others before their own. We are here to serve others. Teammates serve one another and all of Mage serves our community and customers. We play on the same team as our customers; their goals are our goals. We don’t win unless our customers win.

Give people power-ups

power-up

is an object that provides extra abilities or a boost in strength to an individual. Here at Mage, we build each other up by encouraging one another through frequent, honest, and immediate feedback — all with positive intent. We see the unlimited potential in our fellow teammates and in our customers. This is why we believe wholeheartedly in their success and do everything within our power to ensure that.

Think bold and lead the charge

We put the needs of our customers above everything else. If something can be improved for their sake, it’s everyone’s responsibility at Mage to move mountains to make it happen. This is why we don’t let our role define our responsibilities. Everyone at Mage has the authority and power to help customers succeed no matter what.

Deliver magical experiences

Existing AI tools are complex, have a huge learning curve, and are built for experts who are already familiar with AI. In order for us to serve our customers effectively, we simplify every aspect of our work. This enables us to design and create intuitive products that are simple to use and more powerful than anything that exists. We imbue our product with craftsmanship that has been mastered from over a decade of constant practice.

Create compounding forces

Mage operates on an infinite time horizon. This means we prioritize our customer’s success over short term gains. Mage is equipped with resources to go the distance, enabling us to focus on doing what is right for our customers even at our own expense. Here at Mage, we invest in creating flywheels. This means we don’t take shortcuts and we invest a great deal of effort into doing things that have exponential impact.

Level up stats

Our team has an insatiable hunger to actively challenge the status quo. That is why we question all assumptions, have strong opinions held lightly, and never stay complacent. Our curiosity enables us to learn and grow as individuals and as an organization. This is how we continue to evolve and deliver the best results for our customers.

A glimpse of eternity

The future with Mage

True to our core virtue of creating compounding forces and operating on an infinite time horizon, we’ve casted a vision at least 20 years into the future. Here are our writings sent back in time for you to read:

“AI application development evolved beyond a single function and team. Today, AI is created, improved, and applied through the efforts of

Cross-Functional Data Augmentation

. Individuals from different backgrounds, fields, expertise, teams, and functions develop AI through a collaborative process. Mage’s platform has revolutionized the way teams operate and how companies deploy AI.”

“Mage has enabled a new generation of companies:

AI-Centric Organizations

. This new breed of entities are designed and structured from the ground up to orient their internal teams, processes, product, and data to maximize the power of AI.”

“The

Intelligence Network

is a community equipped with the knowledge and power of others in the vast network. Every new node in this network will enhance the capabilities and quality of AI for everyone on the platform.”

“We have advanced the creation and application of AI so much that we successfully made

Multi-Domain Operations

easily accessible to all companies. Prior to this, companies were limited to applying AI in the digital realm. Now through Mage, companies can deploy their AI seamlessly into the physical world.”

“Mage has created the

Omni-Verse

, a portal for companies to harness the power of AI to deliver unimaginable experiences to their customers across every reality.”

Closing thoughts

Mage on the 101 billboard

Mage

is currently in private beta, working very closely with early customers to build the most magical product ever. If you need AI but don’t have the tools or expertise, please send us a message and we would love to chat!

Hang out with us

Join our community and chat about startups, AI/ML, and product development.

Like what you see? Join the team.

Mage is making AI and ML accessible to product developers. Join us and build beautiful and intuitive devtools.

Want to give us feedback or ask questions?

Please chat with us live by joining our Discord channel or send us an email.