---
title: 'Internship - MScAC 2027 at Boson AI'
canonical: 'https://feeny.ai/job/internship-mscac-2027-boson-ai-toronto-px5e22baba9v'
type: 'job'
last_seen: '2026-10-30'
---

# Internship - MScAC 2027 at Boson AI

- **Company:** Boson AI
- **Location:** Toronto, Canada
- **Employment:** internship
- **Work type:** hybrid
- **Posted:** 2026-10-23
- **Last confirmed live:** 2026-10-30
- **Apply:** https://jobs.lever.co/bosonai/3b66f322-55eb-4c4b-9b36-028dfcdd85e9

## Job description

## About The Role

Boson AI builds real-time speech and multimodal AI that makes talking to AI feel like talking to people. Our Higgs models listen, speak, and see. Higgs Realtime is a speech-to-speech API with tool calling and support for 100+ languages, and it is drop-in compatible with the OpenAI Realtime API. Higgs TTS 3 is an expressive, open-weight text-to-speech model with zero-shot voice cloning. Higgs Avatar turns a single image into a streaming, talking video, and Higgs VoiceOver brings directable AI voice-over into video editors. We are a team of 40+ researchers and engineers, and we train our models on our own cluster of 900 GPUs.

Eligibility: this internship is only open to students enrolled in the Master of Science in Applied Computing (MScAC) program at the University of Toronto. Applications from anyone who is not an MScAC student at the University of Toronto will be automatically declined.

We are hiring MScAC students for the program's 8-month applied research internship, starting in May 2027; an earlier start is possible. You will join one of the projects below and work with a Boson AI mentor on problems that feed directly into our models and products. The internship is based in our Toronto office on a hybrid schedule.

Research Projects

- Audio data extraction and annotation. Our speech models learn from very large amounts of audio in 100+ languages. You will build pipelines that collect, segment, transcribe and annotate audio with as little manual engineering as possible, using unsupervised and semi-supervised methods, and you will close the loop so that annotation and model training improve each other (see NVIDIA Granary, [arXiv:2505.13404](https://arxiv.org/abs/2505.13404)). This suits you if you enjoy scripting, experiments and large datasets, and have a working knowledge of statistics and machine learning.
- Audio evaluation and benchmarking. Good speech is more than correct pronunciation. It has to fit the context, emotion and style, sound natural in a conversation, and match the voice it clones. You will design benchmarks and automatic metrics for expressive, multilingual and conversational speech, for voice cloning, and for speech recognition under the conditions real voice agents face (see our EmergentTTS-Eval, [arXiv:2505.23009](https://arxiv.org/abs/2505.23009), and Back to Basics: Revisiting ASR in the Age of Voice Agents, [arXiv:2603.25727](https://arxiv.org/abs/2603.25727)), and use them both to evaluate our models and as signals for training them. This suits you if you like experimentation, have a good ear, ideally speak more than one language, and are comfortable automating evaluation in Python.
- Video generation and computer vision. Higgs Avatar turns a single image into a streaming, talking video. You will work on the data pipeline behind it (collection, multi-stage filtering and synthesis) and on the models: audio-driven video generation with accurate lip sync and natural expressions, distilled diffusion models that run in real time, and computer vision that lets our models see and react to the person they are talking to (see Wan-S2V, [arXiv:2508.18621](https://arxiv.org/abs/2508.18621), and LTX-Video, [arXiv:2501.00103](https://arxiv.org/abs/2501.00103)). This suits you if you have a background in image or video generation (diffusion models, VAEs, GANs, NeRFs) or in computer vision, and are comfortable with large-scale data and model optimization.
- Efficient training. You will make training on our 900-GPU cluster faster and more robust: numerical formats (FP8, FP4, BF16), fused CUDA or Triton kernels for architectures that go beyond standard Transformers, overlapping communication with computation, faster checkpointing and recovery from failures, newer optimizers such as [Muon](https://kellerjordan.github.io/posts/muon/), and stable reinforcement-learning post-training when the training and rollout systems differ (see our Let the Batch Decide, [arXiv:2605.12380](https://arxiv.org/abs/2605.12380)). This suits you if you enjoy systems work and optimization, like writing CUDA or Triton, and want experience training across hundreds of GPUs. Strong coding skills are required.
- Serving and real-time inference. Higgs Realtime responds in under a second, and Higgs Avatar streams live video into calls. You will make our streaming speech and video models faster and cheaper to serve, on our own GPUs and in the cloud, using [SGLang](https://github.com/sgl-project/sglang) or [vLLM](https://github.com/vllm-project/vllm): 8-bit and 4-bit quantization, speculative decoding, kernel and memory-layout optimization, KV-cache management across many concurrent users, and scaling capacity with demand (see our Alignment Drift in Single-Model Speculative Decoding for ASR, [arXiv:2608.12703](https://arxiv.org/abs/2608.12703), and TARQ, [arXiv:2605.27808](https://arxiv.org/abs/2605.27808)). This suits you if you enjoy systems work, have some experience with GPU hardware and CUDA or Triton, and want to work with large numbers of GPUs. Strong coding skills are required.
- Agents: conversational design, theory of mind and reinforcement learning. Our voice agents handle customer support, sales and training calls for businesses. They must follow business rules exactly while sounding natural, keep track of context, and adapt to the person on the other end. Projects include declarative dialog frameworks that separate workflow state from language generation and keep agents on track through multi-step workflows, even after the user interrupts (see our IHBench, [arXiv:2606.19595](https://arxiv.org/abs/2606.19595)); theory of mind, so that agents model the user's state of mind and act on needs the user has implied but not said (see our ProactBench, [arXiv:2605.09228](https://arxiv.org/abs/2605.09228)); and reinforcement learning that improves agents from real call traces (see Flow-GRPO, [arXiv:2510.05592](https://arxiv.org/abs/2510.05592)). This suits you if you are a rigorous logical and systems thinker, know the basics of LLMs and reinforcement learning, and are curious about human behavior. Experience with dialog systems or human-computer interaction is a plus, and so is a love of storytelling or role-playing games.
- Full-duplex conversation, for audio and audio-video. People talk over each other, nod, say "mm-hm" and react while they listen. Turn-based voice AI does none of this. You will work on full-duplex models that listen and speak at the same time, handling interruptions, backchannels and turn-taking, following instructions about when to speak, and keeping up with several people in one conversation: first for speech, then for audio-video, where the avatar also has to look like it is listening and react in real time (see our Instruct-FD, [arXiv:2607.20460](https://arxiv.org/abs/2607.20460), and MSI-Bench, [arXiv:2609.24812](https://arxiv.org/abs/2609.24812)). This suits you if you have a background in speech, sequence modeling or multimodal learning, and want to work on an open research problem with a direct path into our products.

## Minimum Qualifications

- Enrolled in the MScAC program at the University of Toronto. This is mandatory, and applications from anyone else are automatically declined
- Strong programming skills in Python, plus C++, CUDA or Triton for the systems projects
- Solid foundations in machine learning and deep learning, with hands-on experience in PyTorch or JAX
- Careful experimental practice: you design experiments, measure what matters, and report results honestly
- Available for the full 8-month internship and able to work from our Toronto office on a hybrid schedule

## Preferred Qualifications

- Research, coursework or project experience in one of our project areas: speech and audio, video generation, computer vision, LLM agents, reinforcement learning, or ML systems
- Papers, open-source contributions or GitHub projects you can point to
- Experience with large datasets or multi-GPU training
- Fluency in a language besides English. Our models speak 100+ of them

Internship Details

- 8-month applied research internship, starting in May 2027. An earlier start is possible
- Based in our Toronto office on a hybrid schedule
- Compensation: CAD 70,000–100,000 for the 8-month term
- In your application, tell us which one to three projects interest you most

Boson AI is building AI that people can simply talk to. If you want to spend your MScAC internship on speech, video and agents that real people use, we'd love to hear from you. Please include links to code, papers or demos you are proud of.

This posting is for an existing vacancy: a new internship position that we are hiring for in 2027. Boson AI uses artificial intelligence (AI) to help screen and assess applications for this position, and our team reviews the shortlisted candidates.

Boson AI is an equal opportunity employer. We welcome applicants of all backgrounds and do not discriminate on the basis of any characteristic protected by law.

## About Boson AI

## Company Overview
- **One-liner**: Boson AI builds real-time, audio-native AI models and agentic systems for natural, conversational voice interactions with machines.
- **Entity Type**: Private (funding not disclosed)
- **Headquarters**: Santa Clara, California, United States (also Toronto, Canada office)
- **Founded**: 2023
- **Founders**: Dr. Alex Smola (Co-Founder & CEO/CSO), Dr. Mu Li (Co-Founder & CEO/CTO), Yi Zhu (Co-Founder, departed May 2026), Yizhi Liu (Co-Founder)

## Core Business
- **Primary industries**: Conversational AI, Voice AI, Multimodal AI, Enterprise AI infrastructure
- **Target customers**: B2B – enterprises deploying voice agents, real-time customer service, and AI-powered communication workflows
- **Mission**: “Make communication with machines as easy, natural and fun as talking to a human.”

## Products & Services
- **Higgs Realtime**: An end-to-end, audio-native real-time speech-to-speech model for enterprise voice agents. Supports interruptions, code-switching (100+ languages), ~700ms speech-in to speech-out latency, and API compatibility with OpenAI Realtime (change three lines of code). Pricing: $0.0023/min audio in, $0.014/min audio out. [LinkedIn post](https://linkedin.com/company/boson-ai)
- **Higgs Avatar v1**: Real-time conversational avatar foundation model that generates 480p video at 16 FPS from a single image and streaming audio. Designed for real-time digital presence in voice agents. Private preview announced May 2026. [LinkedIn post](https://linkedin.com/company/boson-ai)
- **Agent Platform (Agent OS)**: An internal platform for building, deploying, and orchestrating AI agents (referenced in job postings as “Member of Technical Staff - Agent Platform”).

## Market Standing
- **Valuation**: Not publicly available
- **Key Metric**: Total funding not disclosed; headcount of 32 employees (as of mid-2026)
- **Notable Investors/Partners**: Not publicly listed. Talent sources include Amazon Web Services (AWS), Google, University of Toronto, Vector Institute, indicating strong ties to top AI research and engineering communities.
- **Growth Signals**: Launch of Higgs Realtime (August 2026) and Higgs Avatar v1 (May 2026); active hiring for senior engineering and ML roles; 10,000+ LinkedIn followers with +2.5% monthly growth; technical team comprises 68% of workforce.

## Competitive Advantages
- **Full-stack AI ownership**: Builds own foundation models (audio, avatar, agent orchestration) rather than stitching external components, enabling deep co-design for low latency and natural interaction.
- **Real-time expertise**: Achieves ~700ms speech-in to speech-out and ~125ms barge-in yield, with top benchmarks in tool-calling and interruption recovery.
- **Cost efficiency**: Higgs Realtime pricing ($0.0023/min audio in) is significantly lower than many competitors, targeting enterprise-scale deployment.
- **Research pedigree**: Founders have “almost a century of expertise in AI” (Alex Smola and Mu Li are well-known ML researchers; Smola co-authored seminal work on kernel methods and scalable ML).

## Strategic Focus
- **Real-time conversational AI for enterprise**: Prioritizing production-ready voice agents that can handle interruptions, code-switching, and emotional alignment.
- **Multimodal expansion**: Adding visual presence (avatars) to voice agents to make interactions more natural and human-compatible.
- **Enterprise deployment**: Offering self-hosting options and private previews for large-scale customers.
- **Building the full agentic pipeline**: From data and modeling to training, tuning, and serving – all in-house.

## Why Work Here
- **Culture**: Described as “a diverse group of researchers, engineers, and industry specialists united by a passion for innovation.” Emphasis on building scalable AI that serves millions.
- **Work policy**: Most roles are on-site at Santa Clara HQ or Toronto office. A Site Reliability Engineer role was listed as remote (Toronto). Likely hybrid/on-site for core engineering.
- **Engineering environment**: Deep technical stack (PyTorch, TensorFlow, Kubernetes, NVIDIA, Supermicro, etc.). Opportunity to work on frontier AI models and real-time systems.
- **Growth stage**: Small team (~32 people) with strong research roots; employees have previously worked at AWS, Google, Robinhood, and alumni go to OpenAI, Anthropic, xAI – indicating high-caliber talent and career mobility.
- **Hiring process**: Uses AI tools to assist with resume review and analysis, but final decisions made by humans.

## Sources
1. [boson.ai/about](https://www.boson.ai/about)
2. [boson.ai/about/team](https://www.boson.ai/about/team)
3. [linkedin.com/company/boson-ai](https://linkedin.com/company/boson-ai)
4. [jobs.lever.co/bosonai](https://jobs.lever.co/bosonai)

## Other roles at Boson AI

- [Project Manager - Higgs Audio (Toronto)](https://feeny.ai/job/project-manager-higgs-audio-toronto-boson-ai-toronto-b4rav3bzwd8v) — Toronto, Canada
- [Project Manager - Higgs Audio](https://feeny.ai/job/project-manager-higgs-audio-boson-ai-santa-clara-mgx1ebzy6sc3) — Santa Clara, CA
- [Developer Relations Manager](https://feeny.ai/job/developer-relations-manager-boson-ai-santa-clara-qssv2x0z1qxn) — Santa Clara, CA
- [Software Engineer - Platform & Application](https://feeny.ai/job/software-engineer-platform-application-boson-ai-santa-clara-zct1esjbfzcn) — Santa Clara, CA
- [Site Reliability Engineer](https://feeny.ai/job/site-reliability-engineer-boson-ai-toronto-69xyfzggnjaj) — Toronto, Canada
- [Senior Software Engineer - Systems](https://feeny.ai/job/senior-software-engineer-systems-boson-ai-santa-clara-94h0qsa26h84) — Santa Clara, CA
- [Frontend Engineer](https://feeny.ai/job/frontend-engineer-boson-ai-santa-clara-nbsv99agxh8b) — Santa Clara, CA
- [Machine Learning Engineer - Enterprise](https://feeny.ai/job/machine-learning-engineer-enterprise-boson-ai-toronto-cm87w800xvcq) — Toronto, Canada
- [Member of Technical Staff - Agent Platform (Agent OS)](https://feeny.ai/job/member-of-technical-staff-agent-platform-agent-os-boson-ai-santa-clara-95vzvadf5n6c) — Santa Clara, CA