Blog

Engineering

ChatGPT API: Cheaper, Faster, Wordier

Mar 2nd, 2023

People

Why I Joined Scale

Feb 28th, 2023

People

Visions of Color

Feb 27th, 2023

People

Scale's Generative AI Hackathon

Feb 1st, 2023

Product

SAM Introduction

Scale developed an AI powered system to track movements at these sites using EO satellite imagery. This blog post explores some of the implementation and novel concepts around the problem.

Jan 25th, 2023

Engineering

Why is ChatGPT so good?

OpenAI applied a type of instruction fine-tuning called reinforcement learning with human feedback (RLHF) to enhance ChatGPT. In this blog, understand the role RLHF plays in enhancing large language models and how to implement it.

Jan 25th, 2023

Engineering

Meet Claude: Anthropic’s Rival to ChatGPT

ChatGPT has captured LLM headlines and amazed the AI community with its extensive natural language processing capabilities. A new LLM from Anthropic called Claude is competitive with ChatGPT and offers great promise. We evaluate both models head to head and give you our thoughts on how they compare.

Jan 17th, 2023

Company

Scale AI Named a 2022 Cool Vendor by Gartner®

Jan 11th, 2023

Product

Models are Now Available in Rapid

Rather than doing the heavy lifting of setting up a pipeline for model training, transfer learning, hyperparameter tuning and model hosting, the new Models feature in Scale Rapid enables all software engineers to access a trained and deployed model endpoint (accessible via API) in just a few hours.

Dec 20th, 2022

Engineering

How Much Better is OpenAI’s Newest GPT-3 Model?

We evaluate davinci-003 across a range of classification, summarization, and generation tasks. Using Scale Spellbook, the platform for large language model apps, we show where davinci-003 significantly outperforms the prior version and where it still has room to improve.

Nov 30th, 2022

General

Autonomy Table Stakes: DoD’s Data Layer

Nov 30th, 2022

Engineering

Discovering Emerging Short Form Video Trends with Clustering and HitL

Discovering short-form video trends is challenging. In this blog post, we explain how we use an ML + HiTL approach for faster and higher-quality trend detection.

Nov 28th, 2022

People

Veterans Day @ Scale

Scale’s veteran population is nearly 5% of the US based workforce exceeding the percentage of veterans in the tech workforce, and one-third of veteran employees are recently separated.

Nov 10th, 2022

Company

TransformX 2022 Highlights and Recaps

The 2022 TransformX conference may now be behind us, but you can still catch all of our sessions on-demand! Nearly 32,000 virtual and in-person attendees joined the 3-day event, which included almost 80 sessions and 108 speakers. From navigating an AI-enabled future to putting people on Mars with ML, Alexandr Wang and the Scale team were joined by the world's top industry leaders, researchers, and practitioners of AI and Machine Learning.

Nov 4th, 2022

General

Announcing Scale Forge: Available for Early Access

Catalog Forge is the fastest way to create AI-generated product imagery in seconds. Forge is now in Early Access.

Nov 10th, 2022

Company

TransformX 2022: Operationalizing AI

At this year’s event, we felt it was important to shine a light on how AI is going to define the next era of technology and power.

Oct 19th, 2022

People

Scale Out: National Coming Out Day

The celebration started in 1988 on the first anniversary of the second National March on Washington for Lesbian and Gay Rights in October 1987.

Oct 18th, 2022

Product

Top 4 Tools to Build Your ML Model Evaluation Pipeline

Model evaluation is one of the most important prerequisites prior to shipping an ML model. Read our guidance on the top 4 tools for model evalutation.

Oct 11th, 2022

General

Scale AI: The Data Platform for Autonomy

Come talk to us at the Association of the United States Army (AUSA) conference

Oct 3rd, 2022

Product

Finding Your Best Model Is Tricky—Here's How to Do It!

Learn how you can better evaluate machine learning models.

Sep 29th, 2022

Engineering

Why I Joined Scale

Sep 23rd, 2022

General

How to Make Your Own AI-Generated Images

5 tools to assist you in generating beautiful art from text prompts with ML models.

Sep 22nd, 2022

Company

Scale Achieves FedRAMP ‘In Process’ Designation for its AI Data Platform

Reaching this milestone on the path to authorization is invaluable to Scale’s ability to enable federal, state and local governments in the adaptation of AI applications.

Sep 20th, 2022

General

How Nick Beighton Transformed ASOS by Investing in Product Catalog Data

Nick Beighton, former CEO of ASOS, discusses how he transformed the company from 200 million to 4 billion in revenue by investing in product catalog data.

Sep 20th, 2022

General

The 10 Best Public Datasets for Object Detection in 2022

How to warm-start your computer vision project with these 10 robust public datasets prior to fine tuning on data for your use case.

Sep 13th, 2022

General

Text Is the Universal Interface

Roon guest-authors a post on large language models and the future of computing.

Sep 8th, 2022

Engineering

Why I Joined Scale

Sep 7th, 2022

Engineering

Scale's Talented Engineering Team

Meet the engineers who are accelerating the development of AI applications

Sep 6th, 2022

Product

Introducing Scale Validate for Model Testing

We built Validate as a developer-friendly solution both for API-first and UI-focused users. The Validate interfaces and endpoints make it as easy as possible to get started and integrated into existing pipelines. Today, Scale Validate is used for the testing pipelines of leading ML teams in areas including autonomous driving, surveillance, and industrial robotics.

Aug 24th, 2022

Product

How to Handle Long Tail Data Distributions in Machine Learning

Both research and industry agree that picking out the right data to annotate leads to much better model performance, because it corrects for the naturally unequal data distribution.

Aug 23rd, 2022

Engineering

Machine Learning Assisted Image Segmentation

The pixel-level precision required for highly accurate semantic segmentation makes it extremely challenging and time-consuming for humans to annotate. In this blog, we discuss our Autosegment feature to make it easier and faster for our annotators to annotate.

Aug 22nd, 2022

People

Why I Joined Scale

Scale is working on the world's most ambitious AI projects, with the help of top engineers.

Aug 19th, 2022

Product

Building Great Models with Better Data on Nucleus and Studio

Let’s walk through an example of how you can use two of Scale’s products: Scale Nucleus and Scale Studio to identify and fix bad labels leveraging model predictions.

Aug 18th, 2022

Engineering

How We Scaled Our MongoDB Instance Gracefully

With MongoDB, the burden fell on our Infrastructure team to determine how best to distribute our database nodes across our cloud infrastructure. Read on to learn how we split our MongoDB instance into several shards without downtime.

Aug 18th, 2022

Product

Nucleus Simplified—Dataset Management for Everyone

Nucleus is the centralized, scalable, and collaborative control center for your machine learning (ML) model’s training data. Nucleus helps you update missing or erroneous labels, and identify model failure cases so that you can direct your data curation efforts with the express goal of training better models.

Aug 17th, 2022

General

Build vs. Buy: What Makes Sense for Data Annotation?

When it comes to the question of building vs buying for data annotation, there’s no one right answer. In this blog, we explore the factors to keep in mind as you try to solve your own data annotation build vs. buy equation.

Aug 15th, 2022

People

Why I joined Scale

Scale is in an exciting place where growth opportunities for AI organically grow your career too.

Jul 29th, 2022

General

How to Understand and Improve the Quality of Annotated Data

Training data is the lifeblood of machine learning. Learn how to assess the quality of your annotated data and understand methods to systematically improve the quality of your training data.

Jul 27th, 2022

Product

How to Setup Data Annotation Pipelines

Investing in data annotation or data labeling is key to unlocking the power of AI but data annotation is a more challenging problem than most teams realize. In this blog, we provide a step-by-step guide to build quality data pipelines.

Jul 20th, 2022

Engineering

Using GANs to Align Human Metrics

In this blog, we explore how to effectively apply the principles behind generative adversarial networks (GANs) to our human-in-the-loop data annotation processes to deliver the highest quality data to customers.

Jul 18th, 2022

Company

Scale partners with Deloitte to accelerate AI adoption across the U.S. government

With this partnership, Scale will provide services focused on data management, AI modeling and innovative development to all federal agencies

Jun 28th, 2022

People

Why I Joined Scale

Dennis will be leading Finance, Accounting, Corporate Development, and Strategic Partnerships

Jun 14th, 2022

People

Honoring Pride: Scale Out & The Pink Triangle

Every year since 1995, a pink triangle has been constructed at the top of Twin Peaks in San Francisco during the month of June for Pride.

Jun 6th, 2022

People

Veterans@Scale: Reflection and Remembrance on Memorial Day

As Memorial Day nears, Scale U.S. Army veteran Jacob Sheehan shares his experiences with the costs of war

May 27th, 2022

People

Life at Scale

Scale engineer Sasha Harrison details her experience building ML products at Scale, and what she has learned.

May 19th, 2022

Product

How ML Linting Meaningfully Improves Adaptive Document Processing at Scale

We show that using Adaptive ML models to catch common errors meaningfully improves metrics for our customers.

May 9th, 2022

People

Trans Day of Visibility: My Experience at Scale

Scale Operations leader Willow Primack shares her personal experience for International Transgender Day of Visibility.

Mar 31st, 2022

Product

How to Use Scale for Your ML Hackathon Project

Scale recently sponsored Treehacks, Stanford University’s annual hackathon. In this blog, we discuss how to use Scale Rapid for a successful machine learning (ML) hackathon.

Mar 30th, 2022

Engineering

Design Thinking for ML Systems at Scale

At Scale AI, we are always looking to find ways to leverage ML to deliver the best results for our customers. In this blog post, we discuss the importance of understanding operational and business context to ship ML systems that work.

Mar 28th, 2022

Engineering

Easy Transfer Learning with Scale Rapid

In our latest blog, we explore how to implement transfer learning for object detection using Scale Rapid.

Mar 22nd, 2022

General

Shift Ventures Fellow Spotlight: Bon Strout

Bon Strout shares what he learned in his experience at Scale as a Shift Venture Fellow.

Mar 14th, 2022

People

Why I Joined Scale

Scale board member William Hockey details why he believes AI/ML has the fundamental ability to restructure the way that systems and businesses are built.

Mar 9th, 2022

Company

Scale AI providing free datasets to help national security partners gain current insights into Russia-Ukraine conflict

Starting today, Scale will be providing a series of AI-Ready datasets that algorithm developers can use to rapidly train and deploy AI in support of Ukrainian and NATO operations. By providing these datasets at no cost to national security practitioners, we hope to support a diplomatic solution and swift end to this conflict.

Mar 7th, 2022

Engineering

Using CycleGAN to Train Unsupervised ML Models

In this blog, we explore using CycleGAN to anime-fy images with Scale Rapid.

Feb 15th, 2022

Product

Peer-Reviewed Doesn’t Mean Perfect Data

Don’t trust your dataset just because everybody else does. Learn how to use Scale Nucleus with Rapid to improve your dataset strategically, only sending problematic and business-relevant images to be labeled.

Feb 14th, 2022

Product

Comparing Scale Document AI to Traditional OCR

Even "intelligent" OCR falls short of new ML-based approaches to whole-document extraction, where artificial intelligence models perform context-aware, accurate extraction from messy documents.

Feb 10th, 2022

Product

Introducing: Scale Synthetic - Now Available for Early Access

Scale Synthetic is the most efficient way to enhance ML performance with synthetic data that complements real-world datasets. Talk to us to participate in our Early Access program.

Feb 2nd, 2022

People

Field Engineering at Scale | A Day in the Life

Scale’s Field Engineering is a global team of creative engineers who collaborate with clients to understand their challenges and architect solutions. Learn more about what a day in the life of a Field Engineer at Scale looks like.

Jan 24th, 2022

Company

Scale Highlighted in Gartner Inc.’s 2021 AI Hype Cycles

In an exciting announcement to close out the year, Scale has been featured in two of Gartner, Inc.’s 2021 Hype Cycles. We are honored to have been featured in Hype Cycles for both Artificial Intelligence and Data Science and Machine Learning.

Dec 21st, 2021

Engineering

Machine Learning Papers at NeurIPS 2021

The ML team at Scale is proud to contribute to the field of machine learning (ML) through its own research and partnerships with leading Universities such as Oxford University. Read on for an overview of the three papers accepted to NeurIPS 2021.

Dec 8th, 2021

General

Highlights from TransformX 2021: U.S. Government Sessions

In early October, we hosted our second virtual conference: TransformX. Catch up on some key insights and takeaways that the foremost government leaders have on the future of AI.

Nov 30th, 2021

People

Scale Team Spotlight: Veterans at Scale

As we prepare to celebrate Veterans Day, we wanted to share some of our Scale veterans’ stories to honor them and their commitment to our country, and to also offer advice for those navigating the transition from service to civilian life.

Nov 10th, 2021

Engineering

Faster Semantic Segmentation with Machine Learning-Assisted Workflows

This blog post introduces Autosegment, an ML-powered tool that can segment instances within a human-provided box. We have found that we can delegate high-level reasoning to Taskers (the labelers on our platform) and leave basic semantic segmentation to an ML model. This division of labor allowed us to build a 30% faster generalizable and accurate ML labeling tool through a conceptually simple implementation.

Nov 9th, 2021

General

7 Elements of a Well-Thought-Out AI Strategy: Salesforce COO Bret Taylor

In his talk at the Scale TransformX 2021 conference, Taylor discussed the key considerations that businesses of all sizes should keep in mind to ensure that they delight their customers and achieve sustainable business outcomes.

Nov 8th, 2021

General

On Teaching Robots To Perform Complex Tasks with Dr. Fei-Fei Li

In this TransformX session, Dr. Li shares how vision is critical for first perceiving the physical world and then interacting with it. She explores how recent advances in AI research help machines perceive the environment around them and then engage with it, to perform both short-horizon and long-horizon tasks.

Nov 4th, 2021

Company

SiaSearch Acquisition: Delivering the Full Data Engine to Our Customers

Today we're annoucing the acquisistion of SiaSearch to further enhance data management for Scale Nucleus.

Nov 3rd, 2021

General

Dmitri Dolgov of Waymo Discusses Data-centric AI

‘We like to say that we are not building a vehicle, we are building a driver’, says Dmitri Dolgov co-CEO of Waymo. He sat down with Scale CEO Alexandr Wang to discuss Waymo’s approach to a complex machine learning problem.

Nov 3rd, 2021

General

Eric Schmidt Discusses the Geopolitics of AI

Eric Schmidt, a co-founder of Schmidt Futures and a former CEO of Google, discusses how AI will shape our global future. Eric joined Scale AI CEO Alexandr Wang in this fireside chat at TransformX.

Oct 28th, 2021

General

Netflix Explains Recommendations and Personalization

In this TransformX session, Justin Basilico, Director of Machine Learning and Recommender Systems at Netflix describes how ‘everything at Netflix is a recommendation’. He explores recent trends in recommendations and how they are applied for each of Netflix's 200 million users.

Oct 22nd, 2021

General

HuggingFace CEO, Clement Delangue: Are Transformers Becoming the Most Impactful Tech of the Decade?

In this TransformX session, Clement describes the pervasiveness and capabilities of Transformers. He explores how we might ensure they are used in an ethical and transparent manner.

Oct 21st, 2021

Product

Managing 3D Sensor Fusion Data with Scenes in Scale Nucleus

Following on the announcement of Scale Nucleus, rebuilt from the ground up, Nucleus now supports 3D LiDAR point cloud data using the Scenes paradigm to keep data from multiple sensors organized and manageable.

Oct 20th, 2021

General

Microsoft CTO Kevin Scott's Seven Insights on the Future of AI

In this TransformX session, Kevin Scott, CTO of Microsoft joined Scale AI CEO Alexandr Wang to discuss the most impactful recent advances in AI and how we can enable a new wave of innovation by democratizing AI for the benefit of everyone in society.

Oct 19th, 2021

General

Our Commitment to Securing Customer Data: Scale + ISO 27001

We take the trust of our customers and the security of their data seriously, and today, we are proud to announce the availability of our ISO 27001 certificate.

Oct 12th, 2021

General

Highlights From TransformX 2021 Conference

That's a wrap for TransformX 2021 Conference! Over 23,000 registrants attended 60 sessions from 100 of the world's top leaders, researchers, and practitioners of AI and Machine Learning. Here are the top highlights from some of our favorite sessions.

Oct 15th, 2021

Product

Scale Rapid: Now Generally Available

Scale Rapid is the fastest way to production-level quality labels, with no data minimums. Scale Rapid is now Generally Available.

Oct 6th, 2021

Engineering

Human Annotations Remain Indispensable for Developing Deep Learning Models

Research conducted by Scale AI's ML team has found that human annotations remain indispensable for deep learning models. Learn about our team's latest research paper.

Oct 4th, 2021

Engineering

ML & Human Consensus: The Best of Both Worlds

Achieving high-quality training data in taxonomy categorization is a major challenge. In this blog, we discuss how we leverage a Human + ML consensus pipelines to enhance taxonomy categorization.

Sep 27th, 2021

Product

Improving Datasets and Debugging Machine Learning Models with Scale Nucleus, Rebuilt from the Ground Up

Nucleus, rebuilt from the ground up, makes it even easier to assess data quality, identify edge cases, and automatically categorize and tag objects in your training dataset for labeling.

Sep 13th, 2021

General

Our Commitment to Securing Customer Data: Scale + SOC 2 Type II and HIPAA

We take the trust of our customers and the security of their data seriously, and today, we are proud to announce the availability of our System and Organizational Controls (SOC) 2 Type II report, and the achievement of HIPAA compliance.

Sep 1st, 2021

Company

Scale Rapid: Fast Experimentation with Production-Quality Labels

Scale Rapid is the fastest way to production-quality labels with no data minimums, all with full control.

Jul 19th, 2021

General

Scale Converge: The Future of eCommerce & Retail with AI

Last week we hosted Scale Converge, to showcase the latest advances in AI and ML for eCommerce and online marketplaces. We brought together technical leaders to hear from the people and companies leading the eCommerce and retail industry forward. It was an event full of thoughtful conversations that examined the advancements made, the challenges that still lie ahead, and what we can expect in the years to come.

Jun 30th, 2021

Engineering

How to Tame the Long Tail in Machine Learning

It is often difficult and costly to achieve strong performance on the rare edge cases that make up the long tail of data distribution. In this blog post, we’ll take a deeper look at how sophisticated data curation tools can help machine learning teams target their experiments toward taming the long tail.

Jun 29th, 2021

Company

Scale AI and Research Partner MIT Media Lab Analyze Bias in Dermatological Datasets to Improve Diagnoses

Recent studies demonstrate computer vision models can serve as a useful decision support tool in healthcare but darker skin is underrepresented in datasets. The lack of consideration has been shown to lead neural networks to produce large accuracy disparities. This research sought to determine if adding Fitzpatrick labels would allow researchers to assess algorithmic fairness to ensure better performance.

Jun 23rd, 2021

Product

Own Your Own Maps. Introducing: Scale Mapping

Scale Mapping provides customers with the most flexible, scalable, and transparent mapping solution. With Scale Mapping, customers can generate high precision maps for simulations and real-world testing, enhance prediction and motion planning by effectively predicting another agent’s intent, train perception models to live detect map features, and improve navigation and route planning to maximize efficiency.

Jun 21st, 2021

Product

OCR Is Not Enough. Try Scale Document AI.

Scale Document AI is the next-generation approach for intelligent document processing. Its latest capability, Adaptive AI, deploys refined machine learning models for customers who demand high quality and low latency at scale when it comes to document processing.

Jun 17th, 2021

People

Life at Scale

I strongly believe in our mission to democratize and accelerate the development of AI. What excites me is that we're solving problems that are new, that other companies haven't solved before. We're tackling issues that are going to be more and more important as AI becomes more and more prevalent.

Jun 9th, 2021

General

Doing the Data Work to Mitigate Data Cascades

Scale AI was founded on the belief that better data → better AI. In this blog, we aim to outline the downstream impacts of "bad" data and how Scale aims to mitigate these impacts.

Jun 7th, 2021

Company

Scale AI Named to CB Insights’ AI 100 2021 List

Scale AI has been named to [CB Insights’ AI 100](https://www.cbinsights.com/research/report/artificial-intelligence-top-startups/) 2021 ranking of the most innovative private AI startups – highlighting our leading data annotation services.

Apr 7th, 2021

General

Scale Transform: AI Leaders Showcase Research Breakthroughs and Real-World Impact

Today we hosted our inaugural conference, Scale Transform, to showcase the state of AI today, from the latest research breakthroughs to the real-world impact across industries. It was a day full of thoughtful conversations and presentations that examined the strides made in advancing these core resources, the challenges that still lie ahead, and what we can expect in the years to come.

Mar 26th, 2021

Product

Getting Started with Nucleus

An introduction to Scale Nucleus, a dataset management tool that helps machine learning teams improve their models by improving their data. In this article, we show how Nucleus can help debug model failures, trace them back to dataset issues, and make it easy to prioritize which data to label next.

Feb 10th, 2021

Engineering

Making Pytorch Transformer Twice as Fast on Sequence Generation.

At Scale AI, we use Machine Learning models in a wide range of applications to improve the quality of our annotations. In this blog, we discuss some tricks to drastically improve PyTorch Transformer implementation in just a few lines of code.

Dec 17th, 2020

Company

Scale AI’s Series D: Expanding our Team to Empower AI Dreamers

Today we are excited to share several key updates that build on Scale AI’s mission to accelerate the development of AI. Scale has raised $155M in Series D funding at a valuation north of $3.5B led by Tiger Global.

Dec 1st, 2020

Customers

Scaling Menu Transcription Tasks with Scale Document

How Scale AI combined partial automation with human quality control in our Scale Document labeling pipeline to make menu processing and restaurant onboarding faster and smoother.

Nov 16th, 2020

Product

Announcing Our Series A With Accel

We are excited to announce that we have raised a $4.5 million Series A round of funding led by Accel. Along with this funding, Accel’s Daniel Levine has joined Scale’s board. The funding will be used to invest in our rapid growth, expand our offerings, and grow our team.

Jul 16th, 2017

Company

Scale Hackathon

The Scale AI team hosted its first ever Hackathon last month. The hackathon was designed to: 1. Encourage innovative thinking to solve some of Scale’s biggest challenges. 2. Prototype and potentially launch impactful projects that make Scale better in practice. 3. Facilitate long-term cross-functional relationships. 4. Give all Scaliens a creative break from their daily work.

Oct 28th, 2020

Engineering

Quick Start Guide to Customized API Documentation

Following up on the launch of our new API docs, this post provides further details on how we implemented our docs with Next.JS, Tailwind CSS, and ReadMe.

Oct 20th, 2020

Company

Why I Joined Scale AI

I cannot express how excited I am to join the team at Scale AI. To understand why, one needs to look no farther than my background. Over the past 33 years, I’ve dedicated my career to serving our country, keeping our citizens safe, and improving the technology our warfighters and analysts use to protect us. I’ve seen first-hand how artificial intelligence (AI) already has and will continue to change our world.

May 24th, 2021

Engineering

Scale AI Machine Learning Digest - Q3 2020

Machine learning is a field of study with tremendous strides being made through active academic research. The machine learning team at Scale AI regularly hosts reading groups to discuss papers they find interesting. In this blog post, we go over the papers the team has read throughout the quarter and provide insights on how a paper influenced our own work here at Scale AI when relevant.

Oct 7th, 2020

Company

Introducing: Our New API Docs

The latest update to our API documentation provides a better user experience and more consistent experience between our API documentation and the rest of our platform.

Sep 16th, 2020