Latest AI JobsBrowse Jobs
Tavus logo

Conversational Modelling Research Engineer

TavusRemoteEngineering, Product, & Design
Apply on company site
Salary
Not listed
Job type
Full-time
Work mode
Remote
Posted
22 Apr 2026
Status
Open · checked 7d ago

About this role

Role Overview

The Conversational Modelling Research Engineer position at Tavus is a remote core AI role focused on pioneering foundation multimodal conversational models. This opportunity suits an experienced researcher passionate about advancing real-time human simulation and generative artificial intelligence in a fast-moving startup environment.

Tavus operates as a research lab dedicated to creating human-computing interfaces through AI Humans. These real-time simulation models give machines the ability to see, hear, and respond with visual realism for meaningful face-to-face interactions. The Series B company receives backing from investors such as Sequoia Capital, Y Combinator, and Scale Venture Partners.

Key Responsibilities

Research Large Multimodal Models specifically for conversational avatars, including talking-heads and neural avatars.
Create methods to handle verbal and non-verbal conversational elements while controlling avatar behavior in real time with minimal latency.
Test fine-tuning, conditioning, and adaptation techniques to improve audio-visual multimodal models for specific tasks, expressions, and controllability.
Collaborate directly with the Applied ML team to transition research prototypes into live production environments.
Monitor cutting-edge industry advancements and contribute to defining future technological developments.

Requirements

Hold a PhD, be near completion of a PhD, or possess equivalent research experience in a relevant technical field.
Bring practical experience working with Large Multimodal Models alongside a solid background in generative language models.
Demonstrate past work fine-tuning or adapting Vision-Language Models for control, conditioning, or downstream tasks.
Exhibit a strong foundation in foundation models and deep learning principles.
Possess robust PyTorch skills and comfort constructing deep learning pipelines.

Nice to Have

Familiarity with large-scale model optimization and training procedures.
Background in duplex-conversational models.
Comprehensive knowledge of generative AI across multiple modalities.
Familiarity with software development best practices.
An adaptable, experimental mindset capable of bridging research and engineering.
Academic publications at major venues such as NeurIPS, ICLR, CVPR, ICCV, EMNLP, or COLING.

Skills and Tools

PyTorch
Large Multimodal Models (LMMs)
Generative and Foundation Models
Vision-Language Models (VLMs)
Deep Learning Pipelines
Audio-Visual Multimodal Models

About the Company

Tavus is a research lab building AI Humans to close the gap between people and machines through real-time human simulation. The company enables individuals, developers, and enterprises to build empathetic AI agents available 24/7 in any language. Preferred locations include a hybrid setup in San Francisco or London, with remote options available in the U.S. or Europe for exceptional candidates.

How to Apply

Interested candidates should visit the official Tavus careers page to submit an application for this role.

Skills and tech stack

PyTorch

Questions candidates ask

What is the primary focus of the Conversational Modelling Research Engineer role?

The role focuses on researching Large Multimodal Models for conversational avatars, developing real-time low-latency behavioral controls, and moving research prototypes into production.

Where is the position located?

The position offers remote flexibility within the U.S. or Europe, with preferred hybrid locations in San Francisco or London.

What technical skills are required for this role?

Candidates need strong PyTorch skills, a background in deep learning and foundation models, practical experience with Large Multimodal Models, and familiarity with fine-tuning Vision-Language Models.

What qualifications are expected regarding education?

Applicants should have a PhD, be near completion of a PhD, or possess equivalent research experience in a relevant field.

What are some preferred nice-to-have qualifications?

Preferred bonuses include large-scale model training knowledge, duplex-conversational model experience, broader generative AI understanding, and research publications at top venues like NeurIPS, CVPR, or ICLR.

Ready to apply?

The application is on the official Tavus careers page.

Apply on company site