Inworld AI Tech Stack

Realtime AI infrastructure for interactive agents and voice applications

Software Development Mountain View, California 51–200 employees Founded 2021 Privately Held

Inworld AI builds realtime generative models and infrastructure for conversational agents, with particular depth in voice AI. The tech stack reveals a platform-as-a-service architecture: Node.js + Python backend, Unreal Engine + Unity for client integration, PyTorch for model work, and orchestration via Docker + Kubernetes. Active projects span TTS APIs, model routing, billing systems, and CI/CD automation. The pain-point list (streaming concurrency, low-latency voice serving, interop between C++ and Node.js) matches the engineering-heavy hiring mix and signals a company scaling realtime inference at the infrastructure layer, not just fine-tuning off-the-shelf models.

Tech Stack 39 technologies

Core StackNode.js Playwright Python C++ C# GitHub Actions Jenkins JavaScript TypeScript Go Docker Kubernetes React Next.js Tailwind CSS n8n Zapier Retool PyTorch Terraform Unreal Engine Unity npm LinkedIn X Meta YouTube Material-UI Lemlist Outreach+9 more

What Inworld AI Is Building

◆Challenges

Scaling ai products
Zero to one
Interop between c++ and nodejs
Streaming concurrency performance challenges
Improving test stability
Realtime performance
Shipping quickly
Realtime online
Scaling low-latency voice infrastructure
Cost-effective voice model serving

▲Active Projects

Dynamic a/b experiments
Consumer ai platform
Tts api
Api-based model services
Interactive demos and playgrounds
Web ui testing
System-wide billing
Inworld runtime
Ci/cd pipelines for applications and infrastructure deployments
Graph and node testing

Hiring Activity

Decelerating20 roles · 5 in 30d

Department

Engineering

Research

Marketing

Sales

Design

Product

Seniority

Staff

Lead

Senior

Principal

Mid

Company intelligence

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

Get started free

About Inworld AI

Inworld AI provides realtime generative models and agent infrastructure for developers building interactive AI applications—companions, educational assistants, health tools, and enterprise agents. The platform includes voice AI models, intelligent model routing and optimization, and an Agent Runtime designed to handle millions of concurrent users. Founded in 2021 by former DeepMind and Google (Dialogflow) engineers, the company operates as a research-led product team of 51–200 employees, headquartered in Mountain View. Hiring spans engineering, research, and go-to-market roles across the US, Canada, Switzerland, Germany, and Serbia.

HeadquartersMountain View, California

Company Size51–200 employees

Founded2021

Hiring MarketsCanada, United States, Switzerland, Germany, Serbia

Frequently Asked Questions

What tech stack does Inworld AI use?

Node.js, Python, C++, C#, PyTorch (ML), Docker, Kubernetes (orchestration), Unreal Engine, Unity (client SDKs), Next.js + React (web), GitHub Actions + Jenkins (CI/CD), Terraform (infrastructure).

What is Inworld AI working on?

TTS API, voice infrastructure scaling, agent runtime, dynamic A/B experiments, consumer AI platform, API-based model services, interactive demos, system-wide billing, and CI/CD pipeline automation for applications and infrastructure.