Deci AI

What Deci AI does

Deci AI was an Israel-based deep learning software company, founded in 2019 and headquartered in the Tel Aviv / Ramat Gan area, that built a development platform for optimizing, deploying, and running deep learning models efficiently across heterogeneous hardware environments. Its core technology, AutoNAC (Automated Neural Architecture Construction), automatically constructed and refined neural network architectures to maximize inference efficiency on target hardware — spanning edge, cloud, and on-premise deployments — without requiring manual architecture tuning from engineering teams.

The company operated a B2B, enterprise-focused go-to-market, selling on a recurring subscription basis to AI/ML engineering teams across industries (horizontal segmentation). Deci monetized through its platform and optimization engine rather than a per-seat developer license, with marketing concentrated on direct enterprise field sales, content, and industry events. It raised approximately $46 million in equity capital across a 2020 seed/early round led by Emerge Ventures and Square Peg Capital, a 2021 Series A of $21 million led by Insight Partners with Samsung Next and Vintage Investment Partners, and a 2022 Series B of $25 million again led by Insight Partners; additional strategic support came from Intel Ignite and Comcast NBCUniversal LIFT Labs.

In May 2024, Deci AI was acquired by NVIDIA Corporation of Santa Clara, California, and was dissolved as a separate corporate entity. Its AutoNAC technology and engineering team were absorbed into NVIDIA's enterprise AI software stack. As a standalone company Deci no longer operates; all references to its forward business should be treated as historical.

Name

Deci AI

Legal name

Deci AI

Website

https://deci.ai

Company type

Private

Founded year

2019

Operating status

Acquired

Headcount range

51–100 employees

Short description

Deci AI was an Israel-based deep learning platform company whose AutoNAC technology automatically optimized neural network architectures for efficient deployment across diverse hardware. It served enterprise AI/ML teams on a subscription basis before being acquired by NVIDIA in May 2024.

Ownership category

akta.pro rank

Product category

AI Development Platform

NAICS

Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182), Custom Computer Programming Services (541511), Computer Systems Design and Related Services (54151)

SIC

Services-Prepackaged Software (7372), Services-Computer Programming Services (7371)

akta.pro primary industry

AI Compiler, Runtime & Kernel Optimization Software (CUDA/ROCm/XLA, graph compilers) (HDAAAAAI)

akta.pro secondary industry

End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management) (HDAEANAA)

Keywords

• Deep learning optimization
• Neural architecture search
• AI model acceleration
• Machine learning platform
• AI deployment tools

Headquarters

HQ city

Ramat Gan

HQ country

Israel

HQ region

Middle East

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Marketing or Sales, Operations, Infrastructure

Revenue model

1. AI Model Optimization Platform: Platform providing AI model optimization services for enterprises deploying deep learning models across various environments

Go-to-market motion1 record

Distribution channels1 record

Core offering

Deci AI provided a deep learning development platform that enabled developers and enterprises to build, optimize, and deploy ultra-fast AI models on any hardware. Its core offering centered on AutoNAC, a proprietary Automated Neural Architecture Construction technology that automatically constructed and refined neural network architectures to maximize efficiency, performance, and speed across diverse deployment environments including edge, cloud, and on-premise infrastructure.

Product overview

Deci AI offered a deep learning development platform centered around its proprietary AutoNAC (Automated Neural Architecture Construction) technology. The platform provided end-to-end capabilities for developing, optimizing, and deploying deep learning models across various industries. AutoNAC served as the core optimization engine that automatically constructed and refined neural network architectures to maximize efficiency and performance across different hardware environments and use cases. Following its acquisition by NVIDIA in May 2024, Deci was dissolved as a separate entity and integrated into NVIDIA's enterprise AI software offerings.

Differentiator

Problem solved

Functional benefit

Products and services

• Deep Learning Development Platform End-to-end platform for developing, optimizing, and deploying deep learning models. Enables developers and enterprise AI teams to build AI applications across various industries by providing tools for model optimization and efficient inference across heterogeneous hardware environments.
• AutoNAC (Automated Neural Architecture Construction) Proprietary automated neural architecture construction engine that automatically constructs and optimizes neural network architectures to improve deep learning model efficiency, speed, and deployment across various environments and hardware platforms including edge, cloud, and on-premise.

Segments1 record

Ideal customer profiles1 record

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability1 record

Feature1 record

Partnerships

One partnership is on record.

• NVIDIAcoreStrategic or Co-development Partner · 1 May 2024Deci AI was acquired by NVIDIA Corporation in May 2024. Deci was dissolved as a separate corporate entity and its technology is now part of NVIDIA's enterprise software offerings

Scale indicators3 records

Recent moves6 records

Expansion highlights4 records

Direct peers

• Modular AI: Modular builds a unified AI inference and developer platform (Mojo/MAX), positioned as a faster, more efficient alternative to vendor stacks. It competes with Deci's AutoNAC on hardware-agnostic model optimization and inference performance.
• OctoML: OctoML built an automated ML deployment platform using Apache TVM to optimize models across hardware targets, the same core value proposition as Deci's AutoNAC. Both targeted enterprise ML teams seeking hardware-aware model optimization and inference acceleration.
• Neural Magic: Neural Magic offered software for high-performance neural network inference on commodity CPUs and GPUs, competing directly with Deci's model optimization and efficient deployment focus. Both served enterprise AI teams seeking inference cost reduction.

Emerging players

• Anyscale: Anyscale, the commercial steward of Ray, provides a distributed compute platform for scaling AI workloads including model training and inference. It overlaps with Deci's deployment-side focus for enterprise AI/ML teams.
• Determined AI: Determined AI offered an MLOps platform for training and deploying deep learning models, addressing similar enterprise pain points around efficient model development that Deci targeted.
• Seldon: Seldon provides an MLOps platform for deploying, scaling, and monitoring machine learning models in production, targeting the same enterprise AI/ML team use cases around efficient production deployment.

Broad incumbents

• Hugging Face: Hugging Face operates a broad AI model hub plus inference and deployment tooling (Inference Endpoints, Optimum) that overlaps with Deci's model optimization and deployment platform for the same enterprise AI developer audience.
• NVIDIA TensorRT: TensorRT is NVIDIA's high-performance deep learning inference optimizer and runtime — effectively the in-house competitor to AutoNAC within the same vendor stack that ultimately acquired Deci.
• Qualcomm AI Hub: Qualcomm's AI Hub provides hardware-aware model optimization and deployment for edge AI, overlapping with Deci's positioning of AutoNAC for optimizing models across different deployment environments including edge.

Others

• Apache TVM: TVM is an open-source deep learning compiler stack for optimizing models across hardware backends. It is the underlying infrastructure competitor that both OctoML and parts of the broader ML optimization ecosystem (including Deci's category) were built around.

Market position

Strengths4 records

Weaknesses4 records

Competitive moat3 records

Key risks5 records

Key highlights5 records

Customer concentration

• Xx.com/deci_ai

Revenue estimate

Valuation estimate

Number of profiles

Profiles5 records

Funding overview

Funding rounds5 records

Investors12 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Deci AI