The Prime 5 Knowledge Labeling Companies Based on Everest Group

(Picture courtesy Centific)

The method of annotating and labeling knowledge is vital for supervised studying duties, similar to coaching a big language mannequin (LLM) and different sorts of machine studying fashions. Nevertheless, the necessity for human cognition and enter is a limiting issue on the quantity of knowledge that may be ready. In consequence, there’s appreciable demand for software program that may assist streamline the info labeling and annotation workflow, in addition to for third events that may do the labeling work on an outsourcing contract. Everest Group just lately ranked the highest suppliers in these booming areas.

Everest Group, a Dallas, Texas-based IT analyst agency, analyzed 19 software program distributors and outsourcing suppliers in its Knowledge Annotation and Labeling (DAL) Options for AI/ML PEAK Matrix Evaluation 2024 report. Based on Everest analysts, enterprises primarily worth the velocity at which DAL suppliers can ship the products in addition to the ensuing high quality of the labeled or annotated knowledge.

“They prioritize suppliers that emphasize relationship-building, cost-effectiveness, agility, and a steadfast dedication to ship tangible enterprise influence and RoI all through their transformation journey,” the Everest analysts write within the report. “Outfitted with educated staff and strong annotation platforms, these suppliers effectively information enterprises by the DAL panorama.”

(Picture courtesy Everest Group)

The Matrix ranks every supplier’s market influence towards their imaginative and prescient and functionality, and 5 suppliers made it to the height and are thought of leaders within the house. Listed here are the highest 5, based on their rating.

1. Appen

Appen is the king of the mountain within the DAL house, based on Everest’s report, with superior scores in each MATRIX axes. That’s not stunning, because the Sydney, Australia-based firm has been at this sport for almost 30 years.

The corporate, which is publicly traded and reported $273 million in income final yr, has developed a well-regarded DAL platform and has additionally established DAL outsourcing companies with operations within the US, China, and the Philippines.

Based on Appen, greater than 50-million particular person hours have been spent on its DAL platform, and it has been utilized in greater than 20,000 initiatives, encompassing 10 billion items of knowledge. Greater than 80% of the main LLM builders are Appen customers, the corporate claims, and it has accomplished greater than 100 million LLM knowledge components.

“Appen is devoted to offering prospects with high-quality, reliable knowledge that energy the world’s main AI fashions at scale,” Appen CEO Ryan Kolln stated in a press launch. “With this new accolade, Appen is acknowledged as a cutting-edge market chief within the AI knowledge house.”

2. TELUS Worldwide

TELUS gives a DAL platform (picture courtesy TELUS Worldwide)

The second ranked DAL supplier in Everest’s report is TELUS Worldwide, the Vancouver, Canada-based IT companies large. Along with digital transformation and IT lifecycle companies, TELUS additionally gives knowledge annotation to corporations around the globe.

TELUS bolstered its knowledge annotation enterprise in 2021 with the acquisition of the AI division of Lionbridge. At the moment, TELUS gives a DAL platform that helps a broad vary of knowledge, together with video, nonetheless photos, textual content, sensor, audio, and geo.

Along with software program, TELUS gives an AI Neighborhood that’s composed of greater than 1 million annotation and labelers around the globe. Its knowledge companies run the gamut from knowledge assortment and creation to annotation and validation.

3. Centific

Third place in Everest’s MATRIX goes to Centific, a Redmond, Washington-based firm specializing in offering a variety of companies to facilitate AI, together with knowledge annotation and labeling.

Centific gives the companies of its “domain-segmented annotation groups” that work with the corporate’s customized annotation platform. The corporate, which has operations in India and China, makes a speciality of serving to prospects to arrange knowledge in LLMs, pc imaginative and prescient, speech, search relevance, maps, augmented driving, and augmented actuality/digital actuality (AR/VR).

Along with knowledge annotation, Centific has “many years of expertise” working in knowledge assortment within the LLMs/NLP, pc imaginative and prescient, speech and AR/VR house. It additionally gives skilled experience in reinforcement studying from human suggestions (RLHF), in addition to AI crimson teaming to assist tamp down on LLM hallucinations.

Lastly, Centific can also be a knowledge vendor. The corporate says it has billions of off-the-shelf datasets accessible, starting from name middle audio and stay assembly movies to optical character recognition (OCR) photos and Korean cellphone calls.

4. (Tied) TaskUs

Tied for fourth place on the Everest DAL Options chief board is TaskUs, a enterprise course of outsourcing (BPO) and digital options supplier primarily based in New Braunfels, Texas.

Based in 2008, TaskUs gives a variety of BPO companies, together with name middle operations and content material moderation by its world workforce of 47,000 workers and gig contractors, a lot of whom are primarily based within the Philippines. The corporate went public in 2021, and reported $924 million in revenues final yr.

TaskUs additionally gives knowledge labeling companies for LLM, pc imaginative and prescient, video, and audio. The corporate claims to have greater than 15 years of expertise with knowledge labeling by way of a workforce that has touched 100,000 area consultants in 30 languages.

The corporate touts a human-in-the-loop (RLHF) strategy to creating AI fashions. Along with gathering and labeling knowledge, TaskUs can present knowledge science experience, the whole lot “from preliminary mannequin coaching to steady upkeep and optimization,” the corporate says.

4. (Tied) Akkodis

Additionally tied for fourth is Akkodis, a various engineering firm primarily based in Switzerland that gives a variety of digital companies to purchasers in automotive, aerospace, power, banking, manufacturing, life sciences, healthcare, and IT.

Akkodis, which has €4 billion in annual income and workers greater than 50,000 staff, touts options in huge knowledge, analytics, AI and ML, and robotic course of automation. The corporate can also be shifting into generative AI and copilots.

Whereas co-pilots and GenAI provide great alternative, the corporate says that “there’s much more goes on beneath the floor, and realizing in terms of AI, good knowledge is 80% of the work.”

Akkodis ranked larger than TaskUs on the imaginative and prescient and functionality axis, whereas TaskUs ranked about the identical quantity larger than Akkodis available on the market influence axis, which makes them primarily tied.

Remainder of the Discipline

Everest broke the remainder of the sphere into two teams, together with “main contenders” and “aspirants.”

The foremost contenders embrace iMerit of Kolkata, India, CloudFactory of Durham, North Carolina; NextWealth of London; Innodata of Hackensack, New Jersey; FiveS Digital of Udaipur, India; Sama of San Francisco, California; LXT.AI of Mississauga, Canada; Cogito Tech of Levittown, New York; and Clickworker of Essen, Germany.

Within the aspirants division, Everest lists Digital Divide Knowledge of New York Metropolis; Innominds of San Jose, California; Affect Enterprises of Houston, Texas; and DesiCrew of Chennai, India.

Associated Objects:

Higher Machine Studying Calls for Higher Knowledge Labeling

Knowledge At Extra Than Half Of Firms Will Not Be AI-Prepared By The Finish of 2024

OpenAI Outsourced Knowledge Labeling to Kenyan Staff Incomes Lower than $2 Per Hour: TIME Report