jobs in the wild.

3-5 years experience, no salary bar for right candidate, hsr layout bengaluru we're a small team building models maximizing speech intelligence per unit of energy. we're deep in R&D mode, and the work is as hard as it gets: training models, pushing latency limits, and figuring out things that haven't been figured out yet. we need a founding researcher who can own the science. you'll be the first research hire, which means you'll set the culture, the standards, and the direction - not just execute on someone else's roadmap. Within a few months, you'll be building and leading a small team under you. What we need: hands-on experience training or fine-tuning voice/speech models - tts, asr, vad etc or related areas someone who's shipped models into real products, not just kept them in notebooks comfort working across the stack - from research experiments to inference pipelines enough judgment to know when to build vs. borrow, and when to publish vs. ship what we don't need: someone who's only ever called an API and called it "AI work" a researcher who can't think about latency, cost, or users someone who needs a fully defined problem before they can start to apply, fill out the form in the first comment. research paper, model card, write-ups, demo - something we can actually evaluate. tell us what you built, why it was hard, and what you'd do differently. we'll reach out over email to candidates whose work speaks for itself. thanks :)