get hired at
Hone
AI Engines that own business outcomes, not tasks
- —
- BPFIVSMX+4
- San Francisco
- AI
Open roles at Hone
6 positions we're tracking, re-checked daily
- apply ↗
Member of Technical Staff, Agent Deployments (SF)
san francisco, usaSeniorfirst seen today - apply ↗
Member of Technical Staff, Research Engineering (SF)
san francisco, usaSeniorfirst seen today - apply ↗
Member of Technical Staff, Infrastructure (SF)
san francisco, usaSeniorfirst seen today - apply ↗
Member of Technical Staff, Product Engineering (SF)
san francisco, usaSeniorfirst seen today - apply ↗
Founding Designer (SF)
san francisco, usaMid-levelfirst seen today - apply ↗
Member of Business Staff (SF)
san francisco, usaSeniorfirst seen today
Know when Hone is hiring before anyone else
A role stays uncontested for about four days. Here's the window, and where we put you in it.
From $5.99/week, cancel any time.
- live
- +2h+2hrole spotted & verified1
- +4h+4hyou get the alert1
- day 3day 3you've applied~8
- day 7+day 7+hits the job boards250+
What Hone is building
Hone builds what it calls Engines: persistent AI systems that are handed a business outcome rather than a prompt. The site draws the line explicitly. Chatbots answer questions, agents complete tasks, Engines own the outcome. You give one a role such as outbound rep, procurement coordinator, or credit-risk reviewer, plus goals and guardrails, and it runs for weeks or months against that goal. The interesting part is how an Engine gets competent. Per Index's investment memo, it explores the company's systems, studies how past work was actually done, and writes its own skills and tools. Before it goes live it is replayed against historical decisions in simulation, so you can see what it would have done on old cases before it touches new ones. It runs inside Slack, Teams, or email, keeps organizational memory, logs every action for audit, and asks for human approval on sensitive moves. One Engine is shared by a whole team, so a correction from one person improves it for everyone. The example Engines on the site are all operational back-office loops: inbound lead conversion, vendor management, credit risk, product adoption, customer commitments. At Cognition, Hone is running go-to-market operations. This is the Devin playbook (a planning-and-execution harness for long-horizon work) pointed at the rest of the company instead of the codebase.
Why this matters
Long-horizon business work is exactly where current agents fall apart. CEO-Bench (Chen, Narasimhan & Liu, arXiv 2606.18543, June 2026) had frontier models run a simulated startup for 500 days: only three models finished above the $1M starting balance, and every model evaluated scored below a simple rule-based baseline. The strongest agents were the ones that wrote their own forecasting code for churn, billing timing, and cash. That is close to Hone's design bet. Instead of trusting the model to reason its way through months of noisy operations, the Engine writes its own tooling, is checked against historical decisions before it acts, and escalates the risky calls. Index's thesis is that as models handle longer tasks, value goes to whoever owns the layer between the model and the enterprise. Hone is betting that layer is outcome ownership with an audit trail, not another workflow builder.
Investors: Benchmark (Peter Fenton), Index Ventures (Shardul Shah, Mark Xu), Elad Gil, SV Angel, Angels: Scott Wu (CEO, Cognition)
Hiring outlook
The careers page lists eight live roles, all full-time in the SF office: four Member of Technical Staff tracks (Agent Deployments, Infrastructure, Product Engineering, Research Engineering), a Founding Designer, an Executive Assistant, a Member of Business Staff, and an open 'Make your own role' slot. Index says the team is actively recruiting engineers and operators from Ramp, Midjourney, and Decagon.
Working at Hone
Hone is a AI company based in San Francisco, USA. Working at a AI company at this stage means opportunity to shape your role based on the company stage.
San Francisco, USA is where most Hone positions are located.
How to actually get hired at Hone
Why applying the normal way doesn't work
Hone runs an applicant tracking system, but hiring managers still work referrals first. A cold application to Hone isn't dead, it's just fourth in line behind internal referrals, sourced candidates and recruiter pipelines.
Who to contact at Hone
- Who decides
- the hiring manager
- Best channel
- LinkedIn or direct email
What to show them
Pick one real back-office loop (e.g. inbound lead triage from a public CRM export or a synthetic one) and build a small replay harness. Run an agent's decisions against 50 historical cases, score agreement against what humans actually did, and flag where it should have escalated. Write a one-page postmortem of the disagreements.
A cold email that works at Hone
What Hone screens for
A $60M seed at $285M, five months in, with Benchmark and Index splitting the lead and both partners on the board, is a bet on the team. Fenton and Shah are buying the people who built Devin's harness plus enterprise GTM experience from Harvey and Mercor. Cognition being both an angel (Scott Wu) and a paying customer is the strongest signal: the people closest to Moritz's work are using the product. Hiring is wide open across engineering, design, and ops, and there is a 'make your own role' slot. The founding-designer seat and the Agent Deployments role (the person who makes Engines work at real customers) are where early hires get the most leverage.
Customize your CV for the Hone role. Matching the job description language helps clear ATS filters.
Don't make these mistakes
Moritz built Devin's planning-and-execution harness and then ran the company as Scott Wu's chief of staff, so he has watched long-horizon agents fail more ways than you have. 'I'm passionate about autonomous agents' or a generic LangChain demo is noise. Also skip the 'AI will replace ops teams' framing: Hone's whole pitch is guardrails, simulation against past decisions, and human approval on sensitive actions. Show that you care about an agent being right over months, not that it can take a lot of actions.
Mistakes that kill Hone applications
A recycled CV gets rejected fast at Hone ( people). They notice.
Don't open with what you want. Open with what Hone is dealing with right now — Long-horizon business work is exactly where current agents fall apart, and what you'd do about it.
Most AI applications get ghosted. A day-five follow-up can double your response rate.
Applying to Hone? Get the contact, not the form.
The Hone interview process
We don't yet have verified candidate reports for Hone. What follows is the typical process for a -person AI company, so treat it as a model, not confirmed detail.
Interview stages
- 1
Recruiter Screen
Phone or video · 30 min
What it testsBasic qualification and logisticsUsually run byRecruiter or HR - 2
Hiring Manager Interview
Video call · 45 min
What it testsRole fit and experience deep-diveUsually run byHiring manager - 3
Technical/Functional Round
Video call · 60 min
What it testsSkills assessment and problem-solvingUsually run byTeam members - 4
Final Round
In-person or video · 60 min
What it testsCulture fit and cross-functional alignmentUsually run bySenior leadership
Hone take-home assignment
Hone includes a take-home exercise in their interview process. For AI roles, this typically involves a practical problem that takes 2-4 hours. Focus on clean, working code over premature optimization. They're evaluating how you think and communicate, not just the solution.
Hone interview timeline
Expect 14 days total. Compared to similar AI companies (14 days median), Hone is about average.
Interviewed at Hone?
Tell us how it went: stages, questions, timeline. Takes 90 seconds and it's how this page stays accurate for the next person.
Hone jobs, frequently asked questions
How many jobs does Hone have open?
As of October 11, 2026, Hone has 6 open positions.
Does Hone hire remotely?
Hone doesn't have remote openings at the moment. All roles are in San Francisco, USA.
What roles is Hone hiring for?
Hone is hiring across Other, Engineering, Design. The most recent opening is Member of Technical Staff, Agent Deployments (SF).
How do I apply for a job at Hone?
Use the apply links above, or check our guide to getting hired at Hone.
Does Hone respond to cold emails?
Response rate data for Hone not yet confirmed.
Who is the hiring manager at Hone?
At this size, hiring is usually run by the hiring manager.
How competitive is it to get hired at Hone?
Expect 100-250 applicants in the first two weeks for AI roles at this size. Apply within 72 hours for best odds.
How many rounds is the Hone interview?
4 stages: Recruiter Screen, Hiring Manager Interview, Technical/Functional Round, Final Round.
Is the Hone interview hard?
Expect technical depth and system design, not algorithm trivia. Candidates report Technical Interview as the toughest stage.
Does Hone give a take-home task?
Yes, Hone includes a take-home assignment.
How long does Hone take to get back to you?
Around 14 days across the full process.
What should I prepare for the Hone interview?
technical depth and system design is the priority. Show you can work autonomously — that matters more than algorithms at people.
Where is Hone based?
Hone is headquartered in San Francisco, USA.
Get Hone roles before they're posted
A role stays uncontested for about four days. Here's the window, and where we put you in it.
From $5.99/week, cancel any time.
- live
- +2h+2hrole spotted & verified1
- +4h+4hyou get the alert1
- day 3day 3you've applied~8
- day 7+day 7+hits the job boards250+
Related
More AI companies hiring
Other startups we track in the same space.