Head of Technology - Remote

Chief Technology Officer Kirkland, WA (HQ) — on-site · Type: Full-time · Reports to: Founder & CEO

Hussh (hussh ) is an AI-first research, machine-learning, and product organization building Private Agent One — a private personal AI agent. The human owns their information, holds the keys to it, and shares it only by their consent, through our PCHP (Personal Consent Handshake Protocol). Private Agent One is available now, free for every American; Puppy One, a private computer for the home, is taking reservations; Beside the product company sits a sister business: a permanent-capital asset manager (Fund A / Alphabet 27), run on the same doctrine. Today, the founder is the technology thesis, the architecture review, and the tie-breaking vote on every hard technical question. We are seeking our first Chief Technology Officer — the single authority over what this platform becomes: which systems are worth building, on what architecture, with which models, at what cost — and the leader who makes those decisions true in shipped product.models, post-training, and evaluation; inference on device and in the cloud; data; and security. A personal vault is the highest-value target in consumer computing — its compromise is not an incident, it is the end of the company, and its design has to assume a motivated adversary and a careless moment. In your first 90 days: the technology thesis is written — the bets, the anti-bets, the build/buy/partner calls, each with the evidence that would change your mind; the harness and evaluation architecture is decided, with evals as the gate nothing ships past; Six months in: iMoney has shipped on its date; the engineering organization is structured as the thesis made flesh; The CTO who sets the platform every other seat builds on clears that bar by construction — or the seat was staffed wrong. Own the technology thesis: the multi-year bets on models, architecture, devices, and infrastructure — reviewed against evidence, and revised in writing when the evidence moves Own the agent harness and model strategy: selection, integration, and routing across the frontier model providers (OpenAI, Anthropic, Google, xAI, Meta) and open-weight models — every model behind one evaluation gate, every routing decision measured, no provider privileged by relationship over results Own evaluation as the shipping authority: the benchmarks, regression suites, and human-evaluation design that decide what ships — a product claim our evals cannot back does not ship, and you are the person who says soPCHP as running code, not a document — key management and recovery with no back door, the guarantee that no machine touches a person's information without a consent event that person can inspect and revoke Own security as architecture: threat models for a platform holding a person's whole information life, the review discipline for every system that touches it, and the incident posture — designed in advance, rehearsed, and honest Own inference economics: serving architecture, latency and cost budgets, quantization and distillation for on-device models, and the hybrid routing between the person's devices and our cloud — intelligence that gets cheaper per user as we scale Own the money systems' technical bar: iMoney's ledgers, sweeps, and approval flows — double-entry correctness, idempotency, reconciliation, and the rule that no agent action moves money without the human's explicit approval, enforced in the architecture rather than in a policy Own the platform beneath the portfolio: the shared services, data infrastructure, and relevance systems that every product is built on — one platform, so each product makes the others stronger Own the technical organization: hire the leaders and the bar — founding engineers, research engineers, design engineers — set the review culture and the operating cadence, and represent Hussh's technology as a peer to the frontier labs and the partners whose platforms we build onOwn the technical research agenda: the IIT Kharagpur lab and the Stony Brook AI3 partnership pointed at the problems the platform actually has — personal models, consent systems, on-device intelligence — with results that come home into product Have 15+ years of experience in engineering and technical leadership, with the last 5 years building products in AI — and depth in at least one of: personal information systems, subscriptions, advertising, or relevance and ranking — shipped to real users Have been the final technical authority on a platform that real revenue and real users depended on, and can walk through the two or three decisions — what you chose, what it cost, and what you would do differently Have run product engineering and infrastructure under one hand, or can show why the split failed: you think in systems that include the serving bill Have built production systems on frontier models and under on-device constraints — you know what quantization does to quality, and what a harness owes its models because you have written one Have designed security-critical systems where the asset was other people's trust — keys, payments, health, identity — and your threat models were written before the incidents, not after Write the thesis yourself: your technical documents survive skeptical reading by people who disagree with you, because the reasoning is checkable and the anti-bets are named Time at a frontier AI lab or a category-defining product company in a senior technical leadership seat Both consumer and enterprise platforms in your record Hardware or edge-computing depth: silicon-aware software, devices that ship to homes Money-movement systems — brokerage, payments, or banking rails — where correctness was regulated, audited, and personal A research-community presence: papers, open-source systems, or standards work The role also carries an annual executive bonus tied to company and technology outcomes and a founding equity grant (4-year vest). Comprehensive health, dental, and vision coverage. Employer-matched contributions to Trump Accounts. Support for continuing education, retooling, and retraining — including AI courses, which we actively encourage. Above all: you will work with the latest AI technologies every day, and the expectation — for engineers and non-engineers alike — is to become expert in using them, running evals on your own work until it reaches an A+ standard. Apply here on LinkedIn with your resume and a short note answering one question: tell us about the technology bet you made that the room disagreed with — the bet, the evidence you had at the time, and how it ended. To be considered — the full application: Apply here on LinkedIn, AND email your packet to talent@hushh.ai, subject: [Applications without the email packet are not considered. Tell us about the technology bet you made that the room disagreed with — the bet, the evidence you had at the time, and how it ended. Our evaluations decide what ships. Describe an evaluation system you built or ran that stopped something from shipping — what the gate measured, what it caught, and what changed because it existed. Give a real example where you materially cut cost per user served (or cost per inference) without losing quality — the starting number, the ending number, and the levers you pulled.