List of PyTorch Customers
PyTorch
548 Market St,
San Francisco, 94104, CA,
United States
San Francisco, 94104, CA,
United States
n/a
Professional Services
30
$3M
Since 2010, our global team of researchers has been studying PyTorch customers around the world, aggregating massive amounts of data points that form the basis of our forecast assumptions and perhaps the rise and fall of certain vendors and their products on a quarterly basis.
Each quarter our research team identifies companies that have purchased PyTorch for Apps Development from public (Press Releases, Customer References, Testimonials, Case Studies and Success Stories) and proprietary sources, including the customer size, industry, location, implementation status, partner involvement, LOB Key Stakeholders and related IT decision-makers contact details.
Companies using PyTorch for Apps Development include: Anthropic, a United States based Professional Services organisation with 2500 employees and revenues of $10.00 billion, Lightricks, a Israel based Professional Services organisation with 650 employees and revenues of $150.0 million, Hugging Face, a United States based Professional Services organisation with 500 employees and revenues of $50.0 million, Jasper, a United States based Professional Services organisation with 220 employees and revenues of $45.0 million, Layer 6, a Canada based Professional Services organisation with 100 employees and revenues of $30.0 million and many others.
Contact us if you need a completed and verified list of companies using PyTorch, including the breakdown by industry (21 Verticals), Geography (Region, Country, State, City), Company Size (Revenue, Employees, Asset) and related IT Decision Makers, Key Stakeholders, business and technology executives responsible for the software purchases.
The PyTorch customer wins are being incorporated in our Enterprise Applications Buyer Insight and Technographics Customer Database which has over 100 data fields that detail company usage of software systems and their digital transformation initiatives. Apps Run The World wants to become your No. 1 technographic data source!
Apply Filters For Customers
| Logo | Customer | Industry | Empl. | Revenue | Country | Vendor | Application | Category | When | SI | Insight |
|---|---|---|---|---|---|---|---|---|---|---|---|
 |
Anthropic | Professional Services | 2500 | $10.0B | United States | PyTorch | PyTorch | Apps Development | 2022 | n/a |
In 2022, Anthropic deployed PyTorch to support its trust and safety engineering and oversight workflows, explicitly using PyTorch as part of its Apps Development tooling to build detection and enforcement models. The deployment focused on embedding machine learning into operational monitoring, enabling model behavior detection and automated enforcement actions while surfacing signals to human reviewers.
The implementation used PyTorch to develop abuse detection models, real-time inference components, and model hardening prototypes aligned with safety requirements. Functional capabilities implemented included monitoring systems for API partner behavior, automated enforcement triggers, internal analyst dashboards for manual review, and machine learning pipelines to surface abuse patterns for research teams, leveraging Python, SQL, Scikit-Learn, TensorFlow, and PyTorch as engineering staples.
Operational coverage spanned trust and safety engineering, research, analyst operations, and platform engineering, with systems designed to ingest user reports and telemetry to produce actionable alerts for analysts and researchers. Integrations centered on API partner monitoring and internal tooling for surfacing abuse patterns to research and operations teams, supporting a feedback loop from incident detection to model retraining and hardening.
Governance and workflow changes included instrumenting enforcement workflows to enforce terms of service and acceptable use policies, creating multi-step review paths that combined automated enforcement with manual analyst oversight. The technical approach emphasized lifecycle instrumentation, model retraining inputs from surfaced abuse patterns, and close collaboration between engineering and operational teams to iterate on safety controls.
|
 |
Borealis AI | Professional Services | 150 | $18M | Canada | PyTorch | PyTorch | Apps Development | 2020 | n/a |
In 2020, Borealis AI implemented PyTorch to standardize deep learning research and model development workflows. PyTorch is deployed as the primary framework supporting Borealis AI's Apps Development activities across research and engineering teams.
The deployment emphasizes experimentation and training capabilities, including large scale handling of structured and unstructured datasets, model prototyping, and reproducible training pipelines. Researchers at Borealis AI use PyTorch to develop novel AI solutions, conduct original publishable research, and prepare models for handoff to development teams, with domain-aligned work that includes time series and asynchronous event forecasting when applicable.
Operational coverage includes staff researchers based in Toronto, Waterloo, Vancouver, and Montreal, and the implementation is explicitly aligned with transferring research outputs into the bank's technology capabilities. Governance and workflow changes center on formalizing research to production handoff, aligning machine learning development practices with product and banking experience requirements, and supporting academic publication and conference participation such as NeurIPS, ICLR, ICML, and CVPR.
|
 |
Falkonry | Professional Services | 70 | $11M | United States | PyTorch | PyTorch | Apps Development | 2021 | n/a |
In 2021 Falkonry implemented PyTorch as a core element of its Apps Development toolchain to support time series machine learning and production pipeline orchestration. The engagement concentrated on embedding PyTorch-based model development into a Kubernetes-native pipeline environment at the Sunnyvale based professional services firm, supporting ML engineering and data science workflows.
Falkonry deployed Kubeflow Pipelines and MLRun on Kubernetes to automate and scale pipeline orchestration, with PyTorch used for model training and inference steps. The implementation included a custom featurizer for time series data built with the Scattering Transform, implemented in PyTorch using Kymatio and Python, producing translation invariant representations for classification tasks.
Kubeflow Pipelines and MLRun were instrumented on the Kubernetes platform to orchestrate end to end workflows from feature extraction through model training and packaging, integrating PyTorch model artifacts into automated CI like experiment reproduction and deployment pathways. Operational ownership was centered on the companys ML engineering and data science teams, leveraging containerized workloads and pipeline metadata tracking for reproducibility.
Rollout followed pipeline automation into production orchestration, and the project explicitly reported savings of over 200 work hours monthly through the automated pipeline processes. By embedding PyTorch within this Apps Development architecture, Falkonry standardized time series feature engineering and model lifecycle workflows using the stated technologies.
|
 |
Hugging Face | Professional Services | 500 | $50M | United States | PyTorch | PyTorch | Apps Development | 2022 | n/a |
Hugging Face implemented PyTorch in 2022 as its primary framework to support Apps Development across Machine Learning Engineering and Research functions. The 2022 rollout established PyTorch as a standardized runtime for model experimentation, fine tuning, and reproducible training within Hugging Face engineering teams.
The implementation centered on leveraging PyTorch alongside Hugging Face maintained libraries, specifically the Accelerate library, the PEFT library, and the Trainer API. These components were configured to provide standardized training loops, parameter efficient fine tuning workflows, and device orchestration for single node and multi node GPU training, with the Trainer API used to unify training, evaluation, and checkpointing semantics.
Architecturally PyTorch was embedded into Hugging Face training pipelines and deployment paths to enable consistent model lifecycle handling from research notebooks to production-ready models. The deployment pattern emphasized containerized training jobs on GPU compute, distributed data parallel training using PyTorch primitives, and use of Accelerate for automatic device placement and mixed precision control, aligning framework capabilities with internal CI and model packaging practices.
Governance for the PyTorch implementation focused on repository level standards, code review gates for model and training changes, and upstream collaboration with the PyTorch ecosystem to align library enhancements with company needs. Hugging Face PyTorch Apps Development use is organized around ML Engineering and Research teams, with the implementation shaping development workflows, training configurations, and model packaging conventions.
|
 |
Jasper | Professional Services | 220 | $45M | United States | PyTorch | PyTorch | Apps Development | 2022 | n/a |
In 2022 Jasper implemented PyTorch as a core runtime within an internal MLOps platform, aligning PyTorch with its Apps Development efforts to operationalize fine tuned large language models. The implementation delivered the Jasper Inference Engine, an MLOps platform built on top of the MLRun stack for model registration, inference serving, and monitoring of LLMs fine tuned at Jasper.
The deployment architecture centers on MLRun as the orchestration and registry layer, with PyTorch used as the model runtime and Python as the primary development language. Functional modules explicitly implemented include model registration, inference serving pipelines, and monitoring instrumentation, and Apache Airflow was used for workflow orchestration and automated deployment sequences. Time to deploy a fine tuned LLM into production was reduced to around 1 to 2 days and the environment currently hosts about six production models.
Integrations are explicitly centered on the MLRun stack and Apache Airflow, with PyTorch integrated as the inference framework and model artifacts managed through the platform registry. Operational coverage spans Jaspers machine learning engineering and production inference workloads, focusing on managed deployment and observability for LLMs produced by the company.
Governance and rollout were led by an internal Staff ML Engineer who took the project from ideation to the first milestone and is building the roadmap for subsequent milestones, establishing registration and monitoring workflows as part of operational governance. The narrative reflects a tightly scoped Apps Development implementation where PyTorch functions as the core model execution layer within an MLRun based MLOps topology.
|
 |
|
Professional Services | 100 | $30M | Canada | PyTorch | PyTorch | Apps Development | 2019 | n/a |
|
|
|
|
Professional Services | 650 | $150M | Israel | PyTorch | PyTorch | Apps Development | 2022 | n/a |
|
|
|
|
Professional Services | 100 | $20M | United States | PyTorch | PyTorch | Apps Development | 2023 | n/a |
|
|
|
|
Professional Services | 50 | $9M | Ukraine | PyTorch | PyTorch | Apps Development | 2019 | n/a |
|
|
|
|
Professional Services | 10 | $1M | United States | PyTorch | PyTorch | Apps Development | 2021 | n/a |
|
Showing 1 to 10 of 10 entries
Buyer Intent: Companies Evaluating PyTorch
ARTW Buyer Intent uncovers actionable customer signals, identifying software buyers actively evaluating PyTorch. Gain ongoing access to real-time prospects and uncover hidden opportunities. Companies Actively Evaluating PyTorch for Apps Development include:
- Bds Services Private, a India based Professional Services organization with 600 Employees
Discover Software Buyers actively Evaluating Enterprise Applications
| Logo | Company | Industry | Employees | Revenue | Country | Evaluated |
|---|---|---|---|---|---|---|
 |
Bds Services Private | Professional Services | 600 | $50M | India | 2025-02-03 |
FAQ - APPS RUN THE WORLD PyTorch Coverage
PyTorch is a Apps Development solution from PyTorch.
Companies worldwide use PyTorch, from small firms to large enterprises across 21+ industries.
Organizations such as Anthropic, Lightricks, Hugging Face, Jasper and Layer 6 are recorded users of
PyTorch for Apps Development.
Companies using PyTorch are most concentrated in Professional Services, with adoption spanning over 21 industries.
Companies using PyTorch are most concentrated in United States, Israel and Canada, with adoption tracked across 195 countries worldwide. This global distribution highlights the popularity of PyTorch across Americas, EMEA, and APAC.
Companies using PyTorch range from small businesses with 0-100 employees - 50%, to mid-sized firms with 101-1,000 employees - 40%, large organizations with 1,001-10,000 employees - 10%, and global enterprises with 10,000+ employees - 0%.
Customers of PyTorch include firms across all revenue levels — from $0-100M, to $101M-$1B, $1B-$10B, and $10B+ global corporations.
Contact APPS RUN THE WORLD to access the full verified PyTorch customer database with detailed Firmographics such as industry, geography, revenue, and employee breakdowns as well as key decision makers in charge of Apps Development.