AI‑Native Co‑Founder & CTO — Ship Fast, Scale SaaS

Ascenta HR is a Guided HCM — a human‑capital platform that guides managers and employees in the moment across the entire employee lifecycle, anchored in each company's mission, vision, and values. We're early‑stage and founder‑led, moving from a working proof‑of‑concept to first paying pilots. HR software records what already happened;
Ascenta guides what happens next.

You are the technical founder — you own the technical roadmap and build it. You can plan and execute the long‑term architecture and turn ambiguous, in‑the‑moment requests such as “a customer needs this by tomorrow” into shipped features in hours or days. You hold the long‑term vision in one hand and ship fast in the other, so moving quickly never means chaos.

You'll take Ascenta from a working POC to a real, secure, multi‑tenant SaaS at the pace one AI‑native builder can now move — doing the work of a dozen senior engineers.

This is an AI‑native company; we do not hire people who still write production code by hand. We're looking for someone who orchestrates parallel agentic coding sessions to build highly scalable, reliable production software—at a pace that would take a traditional team months. Our interview is a live build: we hand you a real, loosely‑specified problem and a tight deadline and watch how you go from nothing to something that works.

There's no trick and no single right answer — it simply shows how you actually build, and whether it matches the way we build.

• Agent‑native building. Claude Code and/or Codex is your primary build surface—not an IDE you type in. You ship in days, not weeks.
• Parallel delivery with worktrees. You run multiple agent sessions / git worktrees to build several features at once, not one after another.
• Zero‑to‑one builder. You've taken products from nothing to shipped and in users' hands. You thrive in ambiguity and empty repos, not just mature codebases.
• Full‑stack, modern web. Next.js / React + Type Script, Node, a modern database (MongoDB or Postgres), Vercel‑style deploys. Not backend‑only; not legacy/enterprise‑only stacks.
• Testing from day one. Automated tests + CI/CD from the start; regression across both the API and the UI/UX flow. Comfortable with modern browser / E2E testing (Playwright).
• AI product engineering + cost discipline. LLM integration (Claude/OpenAI), RAG, prompt/token/model‑routing; understands when to use isolated, in‑your‑cloud model access (e.g., AWS Bedrock) for data privacy and compliance vs. calling provider APIs directly.
• Cloud + security/compliance instinct. AWS (including multi‑account org structure for compliant setups), real authentication, multi‑tenant data isolation, PII handling, and a path to SOC 2 — this is HR data.
• Leverage & velocity mindset. You measure shipped value, not lines of code or hours. You plan lean, then build. You challenge “this will take weeks.”
• An optimized, modern workflow. You've reorganized how you work around agents—voice/dictation input (e.g., Wispr Flow), agent memory/skills, MCP—because speed is the job.

• HR / HCM / B2B SaaS experience.
• Prior founder or CTO experience.
• Product and design sense; can own the UX, not just the plumbing.

• Hand‑code everything and treat AI as a bolt‑on rather than the primary surface.
• Are backend‑only, or need a team and process to move.
• Are anchored in legacy/enterprise stacks and “this takes a quarter” estimates.

Founding technical co‑founder with real equity / ownership. Remote, full‑time, all‑in, startup pace. You lead all technical direction and build the platform to first paying customers and beyond.

• What's your primary build surface today—Claude Code, Codex, Cursor, or an IDE you type in? Roughly what % of your code do you write by hand?
• Have you used git worktrees / multiple parallel agent sessions to build several features at once? Describe your setup—or, if not, where you'd expect them to help.
• Walk us through the last thing you shipped solo, end to end—what it was, how long, and the tools.
• How do you keep a growing app from regressing across both the API and the UI/UX flow? What do you use for browser / E2E testing?
• When would you use a managed, in‑your‑cloud model service like AWS Bedrock, and why?