Scale AI: From Shortly Failed Experiments to the ‘Chief’ of Data Quality

AI is playing a significant role in our daily life. As it is widely believed to increase human productivity, AI is in high demand across a variety of industries. Businesses across the economy spend much time and effort collecting data to apply AI for their software tools. However, data must first be labeled before being put into AI models, which is a time-consuming procedure that most businesses perform manually.

Scale AI, a company founded in 2016, aims to leverage automated technologies to complete a large portion of that tagging and identification job. Since its establishment in 2016, the San Francisco-based company has raised more than $600 million in venture funding. Today, Scale AI has aided notable brands in a variety of industries, including Instacart, Etsy, PayPal, Samsung, Toyota, as well as the U.S. Army and Air Force, to build and manage their AI and machine learning models.

From a College Dropout to the World’s Youngest Self-made Billionaire

Before diving into its business model, let’s cast a glimpse over Scale AI’s founding story that has drawn much attention among the public. Astonishingly, the company is run by none other than a 25-year-old, Alexandr Wang and his fellow young tech whizz Lucy Guo. The entrepreneurial duo first met while working at Quora, a QnA platform. Back then, they received investment from widely known Startup accelerator Y Combinator for their venture.

Speaking of Alexandr Wang, he left school when he had just finish freshman year of college. He once again made the mysterious college dropping out story of Bill Gates and Mark Zuckerberg alive. When Wang was just a student at MIT, he showed extreme excitement in artificial intelligence and machine learning, he was interested in how this sphere could change the world.

Whereas MIT was engineering oriented school, people showed less interest in AI. Wang took the plunge to dig into it, the more he explored the AI world, the more bottlenecks about the data he figured out. He realized that it took a lot of effort and resources to make data “intelligent” and suitable for machine learning – and back then, there were no infrastructure or specialized tool to solve this burning issue.

Recognizing the big hole to be solved for AI to ignite its full potential, Wang was inspired to found Scale AI – the data labeling corporation. Now Alexandr Wang becomes the world’s youngest self-made billionaire at the age of 25, the CEO of one of the major AI companies in Silicon Valley with about 700 employees.

The company has been through ups and downs moments in the industry. In the earliest day, Wang and his team admitted getting confused due to the broad field, they literally did not know what kind of data they needed to focus on. Hence, the business decided to try out many options at once before finally serving high-quality training data for AI application companies, especially machine learning – based companies.

This was a tough journey before they could eventually define their missions and focus, many strategies had been conducted and failed short, then changed, but there is one thing that has been consistent, which is Alex’s strong belief system.

This is such a moral study for many startups today. Before rocketing its revenue and acquiring a tremendous growth in the industry, Scale AI made countless testaments to figure out what really works, and more importantly, what doesn’t is not necessarily a bad thing.

A Conceived Factor That Leads to Well-Oiled Machine Alike Operation at Scale

Wang understands that “AI is a supercharger for humans – when AI is better than humans, it makes humans better”

According to Scale employees, the business runs like a well-oiled machine. Together, they’ve accomplished certain tasks that were impossible, and it’s inspiring to see the whole group working cooperatively to perform a difficult task. They obtain a lot of knowledge while working on projects by openly arguing and selecting the best technical solution. One of the finest methods to learn and advance, according to Scale’s engineer, is through seeing and communicating with other engineers.

One of Scale’s guiding principles is “ownership is the job”. To successfully deliver an output, they must take part in product-focused discussions, integrate themselves into operations, and interact with customers directly. Aerin Kim, an engineering manager at Scale AI who oversees a Catalog ML team, claimed that the work and daily activities of a Scale engineer were far more varied than those of any other position she had held before joining, which made for a lively and stimulating working environment.

Scale is a fascinating environment where prospects for AI’s growth organically grow your career as well. Being at Scale enables individuals to work with very visible clients, frequently having an impact on millions of users. They have a framework for decision-making and teamwork according to their credos.

The one that has resonated the most with Vijay Karunamurthy, Head of Engineering in Scale, is “run through walls.’ He rushes to overcome obstacles without being deterred by them. He takes risks, puts ideas to the test, and puts in the necessary effort to find the best solution to a challenging issue. Every member of his staff has demonstrated this credo, as far as he observed.

What Makes This Company Turns into Unicorn?

“Every industry is sitting on huge amounts of data,” Alexandr Wang told Forbes. “Our goal is to help them unlock the potential of the data and supercharge their businesses with AI.”

The team at Scale AI has created a collection of software tools that dramatically cut down the time needed to teach a machine how to analyze and comprehend visual data and lower costs also result from saving time. This company’s product impressively shortens the duration of this procedure by scanning the photos, detecting and labeling the interpreted object, and then asking a human to confirm that the object was correctly named.

The startup is among a handful of established companies that help businesses tackle the data preparation needed to train AI systems, a process that typically requires thousands of annotations with labels. Accordingly, Scale AI is said to be a potential seed growing disruptively in the market as Wang once described that something “like a nuclear reactor for energy and excitement.”

One of their striking products, Scale’s sophisticated LiDAR, video, and picture annotation APIs are beneficial for self-driving, drone, and robotics teams at organizations like Lyft, OpenAI, Zoox, Pinterest, and Airbnb that they can concentrate on creating unique models rather than just labeling data. The business model has also accelerated the process of machine learning in various fields, making it convenient and effective for AI application.

The company is enjoying $100 million in revenue. The data infrastructure for AI company also ranked 22 in the ranking of the top 100 private cloud companies in the world conducted by Forbes in 2022. The company has raised $602.6M going through 7 rounds of funding in total. Scale AI is now valued at over $7 billion after a $325 million funding round in 2021.

Today, there are more than 300 companies, including General Motors, Square, SAP and Flexport, who adopt Scale to glean useful insights from raw data. The leap from a small venture with no more than 100 employees at first to a worldwide scale company of Scale AI has shown that Wang’s vision and determination to follow his consistent belief have been rewarded.

Leading the Tribes: How They Become the Tech Powerhouse behind Other Society-Transformer Brands

We are experiencing state-of-the-art software and technological gadgets in our daily lives, but have we ever wondered about the hidden process and the silent heroes who listen and make them happen? Scale AI is so-called a “tribes’ leader” as it stands behind many businesses and helps them solve complex technical problems along their way of generating the final products or services.

Since it was established, Scale AI has given a helping hand in various spheres, from creating training data for Toyota, handling sensor fusion data for self-driving trucks to delivering seamless management solutions. Using this output from Scale AI, many startups were able to transform it to the input for their social solutions such as faster and safer real estate transactions, self-driving transportations or retail checkout’s enhancement.

On top of that, by providing the ground-truth for businesses’ perception system, Scale AI has cultivated long-lasting partnerships and received many positive testimonials that elevates this brand’s credibility on the market. To demonstrate, the collaboration with other noteworthy brands as given below from different fields is a concrete proof of this rising star.

#1: Standard AI – Retail checkout without ‘embarrassed moment’ recognition

With the bustle and hustle of life, people today get tired of waiting in long checkout lines. Shoppers are also struggling in managing the large amount of goods and eyeing on shoplifters. Standard Cognition is an AI-powered autonomous checkout solution that enables customers to take what they need without having to approach a cashier and assist retailers to manage goods in store effectively using AI applications.

In this collaboration, the San Francisco-based company has helped Standard Cognition handle the volume of annotation tasks using Scale’s platform and the personnel with the necessary expertise to carry out the required annotating duties that Standard Cognition needs.

Although we can’t see Scale expose directly in our daily lives, the company silently brings values to the society, enables shoppers to experience this new paradigm and makes life easier.

#2: Brex – Manage corporate transactions from investors to staff lunch

Brex, a financial services and technology company, is developing an all-in-one finance solution for business. Much of the industry, however, still relies on manual, error-prone workflows including document processing. Brex helps businesses manage and pay their bills in one place simply by uploading a bill or invoice. Back then, one of the biggest issues Brex saw on the market was the reliability of financial software. Even though their solution was applying machine learning, it did not achieve high enough accuracy and low latency.

Scale AI’s involvement has helped Brex generate the output in just seconds using the fine-tuned machine learning models and increases accuracy in as little as two minutes by the optional human-in-the-loop workflow. Collaborating with Scale AI led to faster Brex’s workflows, fewer mistakes, and lower operational costs.

The collaboration went smoothly that Henrique Dubugras, Founder and Co-CEO of Brex regretted not picking Scale AI in the first place. Also, Alexandr Wang said with his appreciation for Brex: “We’ve been with Brex since our early startup days. They’ve helped us grow from a few employees to over 700, and have grown with us.”

#3: Flexport – Follow the journey of your ‘mega package’ from the other half of the globe

Flexport is a technology foundation for international logistics. Before integrating an ML-solution, Flexport struggled to make traditional means of data extraction. As a result, Flexport collaborated with Scale AI in order to more quickly, effectively, and precisely achieve its aim of making international trade simple and accessible for everyone.

By using the ML models, Scale’s Document AI solution met these evolving needs to quickly and easily extract additional fields from unstructured documents without the use of templates. At a time when many people were finding it difficult to obtain any level of insight at all, this solution increased visibility on global trade.

Tom Vu, Head of Data Science & Machine Learning in Flexport gave lots of compliments and claimed that the two brands were continuing to nurture this potential partnership. More document types are being ingested to speed up turnaround times while maintaining accuracy.

The Bold Move from Private Sector to ‘Toughest Buyer’ on the Market: $350 Million in Government Defense Contracts

One noteworthy point is Scale’s Synthetic data can be deployed for the sake of edge-cases in the Army. Additionally, unparalleled autonomous perception can be achieved by fusing and mapping input from diverse sources. This is a potential approach, hence, the government has soon paid attention to the company and shown interest in collaborating. According to Forbes, Scale AI, Wang’s six-year-old San Francisco-based business, has already been awarded three contracts totaling $110 million in 2021 to assist the Air Force and Army of the United States in leveraging AI.

Officially this year, Scale AI is awarded a $249 million-ceiling Blanket Purchasing Agreement issued by Department of Defense’s Joint Artificial Intelligence Center, an organization exploring about AI application in the army, especially when it comes to actual combat. Through the procurement, Scale AI is supposed to readily access all the federal agencies with a view to enhancing the experiences with AI and machine learning capacity.

To date, this transaction has been reckoned as the largest federal contract of the company. Hence, the company decided to put effort to emerge itself in the defense industry. Government and DOD executives have been offered to engage with the company.

“The AI problems that the government has are some of the most interesting and complex and frankly some of the most impactful,” Wang said. The company mission is to ensure a major component of AI models which is the full life cycle of data management for their federal agencies. The company also claims to continue to invest in federal technology. Wang has also shown his deep gratitude for the meaningful business and production business with the DOD.

Bottom Lines

In an interview with Forbes, Business of Business about the future of AI adoption and computers, Wang expected to see Scale AI raising its position prospectively. Personally, I have a strong belief about the growth of technology, especially in the AI industry, it will change our daily lives just as the Internet did, can be even more.

On top of that, the founding story behind Scale AI is admirable. From this moral study, people who harbor to run a startup can ignite the motivation to follow their passion. Especially understanding our own values as soon as possible is really a key factor of success. To me, I could solve many questions which kept swirling around in my head. And with most of the “How to…” questions, the answer is ultimately determination and dedication to passion, I would say.