United's mission is to save people from boring work, starting with SMB accounting. We believe now is the perfect time to bring AI into traditional service industries and fundamentally transform how they operate: services will become a new software.
We’re building Codex/Claude for accountants: AI agents & workflows handle 90%+ of the workload, with real accountants orchestrating the work and owning client relationships.
The role
We are hiring a founding, hands-on engineer to build production AI agents and workflows. These systems operate on real financial data and documents, so correctness, auditability, explainability, and recovery matter as much as model capability.
You will own systems end to end: from prototypes and evaluations through deployment, observability, and iteration with accountants. You will work directly with the founders and operators across our CPA-firm network, with unusual access to live workflows and domain expertise.
What you will own
- Architecting and building AI systems. You decide what context and tools our agents need, how they are structured, how they interact with semi-deterministic workflows, and what each layer produces.
- Defining the boundary between deterministic and nondeterministic logic. You decide which logic should be implemented as fully deterministic code, as a combination of code and AI, or as a full-fledged AI agent.
- Turning signals into action. You observe the AI system at work and speak directly with its users. When its results aren’t good enough, you identify what is missing from the context, pipeline, tools, or architecture and solve it.
- Integrating AI workflows into the product experience. You don’t just improve the metrics of headless agents - you make sure users love working with them.
- Moving the needle. You ensure that your work meaningfully increases human productivity by reliably automating significant portions of their workload.
- Shaping what comes next. Your observations and hands-on experience with moving the needle will directly influence product development and how the product evolves around an agentic paradigm.
What are we looking for
- Strong curiosity about accounting and fintech. To make a meaningful impact in vertical AI, you will need to learn the details of how accounting work is actually done.
- First-principles thinking and the judgment to determine how real work should be distributed among AI agents, AI workflows, and deterministic code.
- Strong software engineering or math/physics fundamentals validating analytical judgement.
- Evidence of taking ambiguous problems from initial idea to measurable outcomes.
- High agency, clear communication, and comfort working directly with domain experts.
- Founder, founding engineer, or early-stage startup experience is especially valuable.
Experience in accounting, finance, or another regulated or document-heavy domain is a plus, not a requirement. No specific school or degree is required. Dropouts are welcomed!
Why join us
- A founding team with a proven track record on both sides of the business: building AI products in this domain and leading large operations at a global scale.
- End-to-end product and technical ownership, with direct access to the founders and users.
- The opportunity to shape the product, agent platform, engineering standards, and technical culture from the ground up.
- The freedom to experiment with AI-native approaches to engineering, with highly flexible token budgets.
- Meaningful equity alongside competitive cash compensation.
- A massive addressable market ($1T+) and a long-term mission that extends well beyond the current market segment.
Location and compensation
This is a full-time role in New York City, five days per week onsite. We work together in person because the fastest path to dependable accounting agents is tight iteration among engineers, founders, and accounting operators.
The final offer will reflect the candidate's abilities, experience, scope, and production-agent track record.
About United
United Accountants is building a national network of independently led CPA firms and bringing the accounting industry into the AI age. We partner with established, owner-led firms, preserving their identity, teams, and client relationships while adding capital, operational expertise, modern technology, and shared resources.
Our mission is to save people from boring work. Agents and automated workflows take on repetitive accounting, tax, and finance work, while accountants stay in control, focus on judgment, and deepen the trusted relationships their clients rely on.
Founding team
Karl Paadam
CEOKarl spent a decade doing the work AI is now automating, at partner level, at the top of the profession. He spent the other decade building and shipping tech.
- Built his first tech company at 16, coded and shipped it himself, exited at 20
- At 26, founding shareholder of a new business at PwC — no clients, no revenue, no team
- Within five years, built it into the largest Tax & Legal practice in Estonia
- Led a team of 500 across 29 countries building and doing acquisitions
- Took a global role at PwC (with 3000 staff across 90 countries); then retired to go back to tech
- Co-founded venture-backed food-tech Yummy, exited to Ruokaboksi in 2024
Peter Potapov
CTOPeter started out as an entrepreneur, then spent a decade building & scaling tech startups - including two accounting products.
- Built a social media business at 16 and exited at 18
- Broadened his tech experience by running a boutique digital agency and joining the founding teams of early-stage startups
- At 22, founded and raised angel funding for Telestat, later exiting through a technology acquisition
- Joined Osome at around $100K ARR and led development of its accounting product
- At 25, promoted to Head of Engineering & ML; scaled the team to 50+ engineers and left as Osome reached $20M ARR
- Joined Kick as its founding Head of Technology. Built its accounting ledger solo from scratch in three months, scaled the engineering team, and later led AI engineering
- During his tenure, Kick grew from $3M to $22M in funding and achieved 30% MoM accounting product growth
Culture
Transparency
Default open — work, numbers, decisions, all visible. Transparency is how trust gets built, and its absence is how trust dies; that’s why trust isn’t a separate value — it’s the product of this one. Public and universal: one company, one culture, and every firm we acquire joins it.
Agency
See a problem, own it, solve it. Freedom to act, accountability for the outcome, nobody waiting to be unblocked. Curiosity lives here — agency without going deep is just motion.
Urgency
A sense of time, not a state of rush. What can be done today doesn’t wait for tomorrow, because early companies die of slow. Urgency never excuses sloppy — speed and quality are both the standard.
Judgment
Know what good and bad look like, and what matters most right now. Think from first principles, go deep to understand the tradeoffs, distinguish between short- and long-term solutions, and hold the work to the highest standard.
