Cerence AI Tech Stack

AI voice and conversational platform for automotive and mobility OEMs

Software Development Burlington, Massachusetts 1,001–5,000 employees Founded 2019 Public Company

Cerence AI builds voice, speech, and generative AI systems embedded in automotive and transportation products, with over 525 million vehicles shipped carrying their technology. The tech stack reveals a dual-track engineering approach: production optimization via ONNX Runtime and TensorRT (model inference on embedded devices), paired with training infrastructure (PyTorch, TensorFlow, GANs) and emerging large-language-model work. An engineering-dominant hiring profile (50 engineers vs. 6 sales) and active projects across ASR, text-to-speech, and dialog AI signal a company scaling AI product breadth while managing the operational complexity of embedding models in automotive hardware.

Tech Stack 97 technologies

Core StackJira Confluence Python PyTorch TensorFlow Java JavaScript TypeScript C++ Linux GANs Crowdstrike Falcon Azure Drata C/C++ ONNX Runtime TensorRT TorchScript BERT VITS HiFi-GAN Android Bluetooth HTTP TCP Android NDK Git JNI JavaScript/TypeScript Android Automotive OS+63 more

AdoptingAndroid JNI WEKA Ceph Slurm Scality Azure

What Cerence AI Is Building

◆Challenges

Maintaining devops pipelines
Reducing response latency
Optimizing cpu memory footprint
Cold-start acceleration
Entering new ai-driven segments
Scaling genai business
Operationalizing ai models at scale
Named entity awareness
Content purity
Daily manual tasks

▲Active Projects

Next-generation text-to-speech system
Voice interaction middleware
Speech enhancement technology qualification for embedded and cloud asr
Hybrid llm solution
Asr and wake-up word model development for automotive and mobility markets
Dialog ai system
Destination entry ai system
Search ai system
State-of-the-art generative ai models
Hybrid voice recognition system

Hiring Activity

Steady85 roles · 30 in 30d

Department

Engineering

Research

Sales

Data

Product

Design

Security

Finance

Seniority

Senior

Principal

Mid

Manager

Intern

Director

Company intelligence

Find more companies like Cerence AI by tech stack, pain points and active projects

Get started free

About Cerence AI

Cerence AI is a public company (NASDAQ: CRNC) that develops voice, speech recognition, and conversational AI experiences for global automakers and transportation OEMs. The platform spans multiple interaction modalities—voice commands, text-to-speech, ASR, and dialog systems—designed to run on embedded automotive hardware and cloud infrastructure. With operations across the United States, Germany, Canada, Japan, Taiwan, United Kingdom, India, and South Korea, the company maintains a 1,001–5,000 person organization focused on advancing AI innovation in the automotive and mobility sectors.

HeadquartersBurlington, Massachusetts

Company Size1,001–5,000 employees

Founded2019

Hiring MarketsGermany, Canada, United States, Japan, Taiwan, United Kingdom, India, South Korea

Frequently Asked Questions

What AI technologies does Cerence use?

PyTorch, TensorFlow, GANs, BERT, VITS, HiFi-GAN for training and modeling; ONNX Runtime and TensorRT for embedded inference optimization; and hybrid LLM solutions. The company is adopting WEKA, Ceph, and Slurm for distributed compute.

What is Cerence working on?

Next-generation text-to-speech, voice interaction middleware, ASR and wake-word detection for automotive, hybrid speech recognition systems, dialog AI, and generative AI models. Key challenges include reducing latency, optimizing CPU memory footprint, and operationalizing AI models at scale.