Hyperbound is seeking a Machine Learning Engineer to own the full lifecycle of models for its revenue activation platform. The role involves fine-tuning and deploying open-source models, building evaluation frameworks, and ensuring production reliability in a fast-paced, onsite environment.
Full description
We’re probably not the right fit for you.
The Reality
We're in the office five days a week in San Francisco. We move fast and expect a lot. We’re most effective when everyone is in the same room, thinking out loud, catching each other's mistakes early.
You'll own your models end to end, training through production. We trust you to handle that. You'll spend most of your time working fine-tuning and running open source models in production, pushing models on-device, and building the evaluation frameworks that tell us whether a change actually worked instead of trusting our gut.
This will probably be one of the hardest jobs you've had. We think it'll also be one of the more exciting ones, and we'll be here right alongside you.
Sound exciting? Good. We’re excited too.
What We're Building
Hyperbound is the Revenue Activation Platform, the agentic operating system for sales. We don't just record what happens on a call, we change what happens next, turning real selling behavior into targeted roleplays, coaching, and workflow changes that make reps better without adding a single layer of management overhead.
Founded in 2023 out of YC (S23), we're 45 people now, Series A, $18M raised total. Last quarter we doubled headcount, revenue nearly doubled with it, and net revenue retention landed above 300%. Teams at LinkedIn, Workday, and Intel run their sales orgs on us.
The Role
You'll build and ship the models underneath Hyperbound's roleplay, scoring, and coaching products: real systems sitting on real sales calls. That means owning the full lifecycle, from training and fine-tuning through getting a model into production and keeping it working once it's there.
You will fine-tune and deploy open source models where they give us more control over cost, latency, or what the model can actually do. Some of it means getting models running on-device, wherever a customer's latency or privacy requirements demand it, with all the tradeoffs around quantization and distillation that come with it. And a good amount of it means building the evaluation frameworks, benchmarks, and regression suites that tell us whether a change actually made things better, before a customer finds out for us.
You'll work closely with the founders and the rest of engineering, and you'll have real input into what we build next, not just how we build it.
The Team
We're assembling a cracked team of builders, operators, and hunters. We've hired people from our competition, former co-founders, and early customers who loved the product.
The engineers you'd work alongside are builders first. Some have started their own companies. Nobody here is precious about their code, and nobody hides behind "that's not my problem," people just see something broken and fix it.
We work hard and we celebrate for real. We’re demoing features we shipped that morning, and the whiteboards get erased and rewritten constantly. You can feel it when a team actually believes in what it's building, and that's very real here. Here's the team in Bali after we raised our Series A:
At the same time, this isn't a mattress-under-your-desk startup: plenty of people here have kids, people leave to make dinner, and they come back the next morning and ship. High standards don't require burnout.
We've doubled in size since this photo was taken a few months ago:
Ownership and Equity
You own your models end to end: the training, the eval, the deployment, and everything that happens after. That cuts both ways, and when it works, that's yours too. Equity here is real, with real secondary opportunities, and we genuinely want the people who build this early to end up meaningfully rewarded for it.
Compensation and Benefits
Comp: $260k-$300k+ based on experience, meaningful equity
Benefits: medical, dental, vision, 401k
Commuter and parking benefits
Unlimited PTO
Free lunch and dinner in the office
The Interview Process
We move fast: an intro call, a technical conversation with the team you'd actually work with, and a final conversation with the founders. We move fast, 1-2 weeks from first conversation to offer.
Equal Opportunity
Hyperbound is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees, and we don't discriminate on the basis of race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability status.
Company information
Hyperbound AI analyzes thousands of your sales calls, uncovers why your top reps win, and builds AI roleplays to help you quickly evolve and reinforce your playbooks as you scale. We pioneered AI Sales Roleplay technology and we're transforming the way mid-market/enterprise revenue teams upskill and prepare their workforce. We provide human-like roleplays across the entire sales cycle with AI buyers customized for your ICP, giving reps more at-bats and leveling up their business acumen.From cold calling to discovery, negotiation, and renewal/upsell/cross-sell scenarios, you can replicate real buyer scenarios, and help your team close more deals. The goal is faster ramp times, more pipeline, higher win rates, increased ACVs, more consistent performance across your team, and more time for your managers.Interested in trying it out today? Start a free demo simulation at hyperbound.ai and contact us at founders@hyperbound.ai for next steps.
