Hands-on Voice AI Software Engineering Internship

Madrona Venture Labs; United States; On-site; USD 16,740 - 33,480; Full time

Job summary

Deepgram is seeking motivated students for Fall 2026 and Summer 2027 cohorts to join the engineering team. You will own a real project end-to-end, working on production-grade voice AI platforms like real-time ASR and TTS, and collaborate closely with research, engineering, and customers.

You’ll prototype with AI tools and contribute to meaningful, shipped code. Ideal candidates are pursuing a CS/engineering degree or equivalent, with coursework or projects in ML, real-time systems, or audio

Qualifications

• You’ve built things: projects, tools, scripts, or automations.
• You reach for AI as a default in learning and building.
• You reason from first principles when something breaks.
• You write and read code in at least one language and pick up new ones quickly.
• You can explain your work clearly and what you'd do differently.
• You strive for better, not just good enough, and you welcome feedback.

Responsibilities

• Design, build, and ship a scoped project end to end with a mentor.
• Contribute to a production codebase across Deepgram’s platforms.
• Explore voice AI, real-time systems, and feedback loops with customers.
• Prototype using Agentic tooling and propose workflow improvements.

Skills

AI mindset

First principles

Coding skills

Clear communication

Feedback oriented

Education

Currently pursuing a degree in computer science, engineering, or a related field

Job description

Deepgram is seeking motivated students for Fall 2026 and Summer 2027 cohorts to join the engineering team. You will own a real project end-to-end, working on production-grade voice AI platforms like real-time ASR and TTS, and collaborate closely with research, engineering, and customers.

You’ll prototype with AI tools and contribute to meaningful, shipped code. Ideal candidates are pursuing a CS/engineering degree or equivalent, with coursework or projects in ML, real-time systems, or audio