Software Engineer - Applied AI
Deepgram · San Francisco, CA
About this role
## Software Engineer - Applied AI ### Company Overview Deepgram is the leading platform underpinning the Voice AI economy, offering real-time APIs for **speech-to-text (STT)**, **text-to-speech (TTS)**, and production-grade **voice agents** at scale. Trusted by **200,000+ developers** and **1,300+ organizations**, Deepgram powers voice offerings “**Powered by Deepgram**” across companies including Twilio, Cloudflare, Sierra, Decagon, Vapi, Daily, Cresta, Granola, and Jack in the Box. ### Operating Rhythm (AI-First) - An **AI-first mindset** is core to how you operate, innovate, and measure performance. - You’re expected to **actively use and experiment** with advanced AI tools (and even build your own). - You’ll **adopt new models quickly**, integrate AI into your workflows, and continuously push what’s possible. - The pace is fast—your day-to-day work will evolve rapidly. ### Team Overview The **Applied AI** team tackles voice AI problems that don’t have off-the-shelf answers. You’ll work directly with enterprise customers and technology partners, often on **zero-to-one** efforts. **Current focus areas include:** - Robust **ASR** in noisy outdoor environments - **Self-service fine-tuning** and model adaptation - **Domain-aware voice agents** for drive-thru and phone ordering - **Agent scoring and evaluation** - Real-time **operational analytics** - On-site **hardware integration** The team is small, works **in person in San Francisco**, ships to production quickly, and stays close to customers. ### Key Goals - Ship production features in live voice agent products (e.g., drive-thru/phone ordering) within your **first quarter** - Own customer problems end-to-end: identify root cause and get fixes into **production** - Build reusable agent modules (e.g., ordering/reservations) to help new customers launch faster - Improve accuracy on real-world menus/catalogs while keeping structured data consistent across locations - Contribute to fine-tuning/model adaptation so customers get accurate results on their vocabulary/domains - Spot product issues early and fix them (or propose solutions) - Run experiments to prove new functionality improves customer outcomes ### Minimum Skills, Knowledge & Capabilities - Experience building and shipping **backend services** that run in production in the cloud - **AI-native** development: modern AI coding tools as a default workflow (not occasional) - Learns quickly and can take on unfamiliar parts of the stack with minimal ramp-up - Proactive problem-solving: recognize issues, propose solutions, and follow through - Experience taking something from **idea to production** (at work or on your own time) - Clear communication; comfortable working directly with teammates and customers - Builds for customer outcomes with a high quality bar and no over-engineering - Comfortable working across languages (stack mainly **Kotlin** and **Python**, with **TypeScript** on the front end) ### Preferred Qualifications (Nice-to-Have) - Hands-on experience training/fine-tuning/adapting models - Experience building evaluation/scoring for models or agents - Experience with **audio**, **speech-to-text**, or **text-to-speech** - Experience as a founder/early startup engineer or taking systems from **zero to one** - Typically **2+ years** of engineering experience (or equivalent demonstrated experience) ### Security / Recruitment Notice Deepgram is aware of impersonators. Legitimate recruiting communication comes from an **
Listing freshness
CronJobs last confirmed this listing 2h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.