Deep learning lifecycle made simple,
with outstanding performance.
For deep learning teams looking to quickly achieve outstanding accuracy & inference performance.
TALK TO AN EXPERTFor deep learning experts looking to meet specific performance goals for highly customized use cases.
TALK TO AN EXPERTFor developers looking to boost performance and shorten development time.
For deep learning teams looking to quickly achieve outstanding accuracy & inference performance.
For deep learning experts looking to meet specific performance goals for highly customized use cases.
Absolutely. The Community plan allows you to quickly start with your own deep learning models or start with a pre-loaded model from our ModelHub. You can upgrade your plan at any time by contacting us.
No. It’s free forever. Knock yourself out.
No. Credit card is not needed.
No limit, go for it.
There is no limit to the numbers of optimizations you can carry out. Keep in mind that if you are in the Community plan, you are limited to using any 5 of those optimized models.
AutoNAC, short for Automated Neural Architecture Construction, is Deci’s proprietary optimization technology. It is a Neural Architecture Search (NAS) algorithm that provides you with end-to-end accuracy-preserving hardware-aware inference acceleration. AutoNAC considers and leverages all components in the inference stack, including compilers, pruning, and quantization. Learn more about our technology. Learn more about the kind of results you can get.
Deci's deep learning development platform enables you to push models from the Deci Lab, optimized or not, directly to one of our inference engines for seamless deployment and ultra-fast serving. Deci offers 2 types of runtime engines: 1. Quickly run inference on your models with Infery, a Python inference runtime engine that simplifies deep learning model inference across multiple frameworks and hardware using only 3 lines of code. 2. Scale your workloads with RTiC (Runtime Inference Container), a high-performance container-based inference server. RTiC fits best with microservice environments and has built-in performance optimizations.
You can upload and optimize as many deep learning models as you want. The Community plan limits you to using any 5 of those optimized models for inference. Once you see how the different optimizations add value to the original model, you can choose which of the optimized models you want to deploy. The Professional plan includes unlimited usage of optimized models. Get in touch to get a quote.
Absolutely. You can invite an unlimited number of co-workers using the ‘Invite’ button within the platform. All of you will be able to collaborate on the same workspace.
Your data’s confidentiality and privacy are our highest priority. We comply with strict information security practices, and are ISO270001 and ISO 27799 compliant. When you optimize your own model, it is uploaded to our server. We never use your model for any purpose other than its optimization on the platform. For the Community and Professional plans, you do not need to upload any data aside from the model itself (e.g., dataset). Alternatively, you can use a pre-loaded model from our ModelHub. For the Enterprise plan, your data will serve as another input to our AutoNAC optimization engine to ensure that you get the full value Deci has to offer. The process can be done entirely on your premises, without the data ever leaving your site. To learn more, please refer to our Privacy Policy.
Absolutely. You can easily integrate Deci’s deep learning acceleration platform using our API access. Read more about our API access.
With Deci’s deep learning development platform, developers can accelerate inference performance by up to 15x, for any hardware, without compromising accuracy; cut up to 80% on compute costs; and reduce time-to-production of models.
SIGN UP FOR FREE
"By collaborating with Deci, we aim to help our customers accelerate AI innovation and deploy AI solutions everywhere using our industry-leading platforms, from data centers to edge systems that accelerate high-throughput inference."
Head of Advanced AI Solutions & Technologies, HPE
“At RingCentral, we strive to provide our customers with the best AI-based experiences. With Deci’s platform, we were able to exceed our deep learning performance goals while shortening our development cycles. Working with Deci allows us to launch superior products faster.”
Senior Vice President R&D, RingCentral
“Deci exceeded our expectations in achieving aggressive targets for reducing latency and memory usage on TPU edge hardware in a test environment, all while preserving BriefCam’s high standard for accuracy.”
Sr. Product Manager, BriefCam
“Deci delivers optimized deep learning inference on Intel processors as highlighted in MLPerf, allowing our customers to meet performance SLAs, reduce cost, decrease time to deployment, and gives them the ability to effectively scale.”
AI Solutions and Sales Director, Intel
"With Deci we were able to get 6.4x higher throughput on our detection model. Its easy-to-use platform allowed us to quickly optimize and benchmark for choosing our best configuration of hardware and parameters."
CTO, UNX Digital by Grupo Prominente
"The classification model I uploaded and integrated using Infery achieved a 33% performance boost, which is very cool for 10 minutes of work!"
Algorithm Developer, Sight Diagnostics
"We are excited to be working with Deci's platform - it provided amazing results and achieved 4.6x acceleration on a model we ran in production and helped us provide faster service to our customers.”
CEO, WSC Sports Technologies
“With Deci, we increased by 2.6x the inference throughput of one of our multiclass classification models running on V100 machines - without losing accuracy. Deci can cut 50% off the inference costs in our cloud deployments worldwide.”
CTO and Co-Founder, Ibex Medical Analytics
Selected as one of the top 100 AI startups in the world
Recognized as a tech innovator for edge AI
Achieved 11.8x acceleration in collaboration with Intel
Named a leader among AI optimization solutions
