Senior Full-Stack Engineer

We are a SEED stage fintech based in SF. Founded by an experienced team with multiple exits and deep knowledge of our target market. Successfully rolling out to a major enterprise customer now, with a robust pipeline to scale the business rapidly.

What we're building

Worthy builds AI-enabled reasoning assistants for financial advisors. Advisors sit on a pile of client financial documents - 1040s, 1099s, K-1s, W-2, brokerage statements, state filings, and the useful advice is hidden, because accurately processing them at scale is slow, tedious work. To solve this, Worthy engineers data pipelines to detect, classify , and extract connected information. Beyond data pipelines, Worthy stack is a combination of traditional ML, LLM's and computational algorithms that generate a range of computed scenarios, today covering a broad scope of tax planning topics. Your work will be at the forefront of utilizing non-deterministic tools to deliver grounded, compliant planning recommendations.

The hard part isn't building a feature. It's presenting accurate, verifiable and repeated information without a deterministic, form driven interface.

Our users love asking questions in natural language, and translating a conversation to an accurate, verifiable and repeated financial analysis is essential. Worthy delivering a wrong number can lead to erroneous advice for a real client, and that can not happen So Worthy engineering is a combination of building accurate, secure, scalable products : extraction pipelines that know when they're unsure, evals to catch regressions before customers do, admin tooling to assist human review, and audit trails for enterprise customers.

The role

Own a product surface end to end - backend, frontend, the AI workflow behind them, and the on-call pager when they break. Concretely, that means:

• Full-stack features work in Go, Python, React, TypeScript, PostgreSQL, and ConnectRPC. You write the service and the screen that uses it.
• AI workflows for document extraction and validation, advisor chat, and the review and eval systems that keep both honest.
• Turning tax rules into product behavior. You won't need to arrive as a tax expert, but you will need to enjoy sitting with a rule until the edge cases are pinned down, then encoding it precisely.
• Internal tools for QA, document reruns, human review, and eval analysis. These are first-class products here, not scripts someone tolerates.
• Production ownership: investigate, fix, ship, then add the guardrail that stops it recurring. Reliability, observability, and deployment safety are part of the job, not a separate team's problem.
• A mind that experiments and tests various options before implementing the most optimal one. Understanding how to extract and drive repeatability from LLM's, understanding the importance of security, auditability, scale and resiliency.

What a good fit looks like

• Senior-level full-stack experience, with deep expertise in building backend systems, data modelling, distributed processing pipelines and enough frontend ability to build a workflow without a handoff.
• You have a deep understanding of LLM capabilities and tradeoffs across model tiers, model routing and orchestration, and decomposing tasks across ensembles of models to optimize for latency, accuracy, and token cost. You know when a problem calls for prompt engineering versus fine-tuning, and can reason about what works in a notebook versus a system that retrieves thousands of documents at scale.
• You treat the harness - tools, context management, verification loops - as a first-class engineering surface, not an afterthought to the model.
• You debug well across layers. Application code, infrastructure, and data all lie in different ways, and you're comfortable in all three.
• You've shipped in a fast-moving environment where requirements arrived half-formed and it was your job to make them concrete.
• Regulated or high-accuracy domains don't feel like a constraint you're fighting.

Helpful but not required: tax, fintech, wealth management, accounting, or compliance background; AWS/GCP production experience; work on evals, prompt governance, RAG or search grounding; open weight models, pruning, tuning; OCR and document extraction; or a track record of building ops and human-review tooling.

This position is based in our San Francisco office, with a minimum of at least three days per week in-person. We offer a comprehensive benefits package to support your well-being, including a company-sponsored retirement plan, flexible health coverage through an Individual Coverage Health Reimbursement Arrangement (ICHRA), and paid vacation.

Equal Opportunity Employer

We are an equal opportunity employer dedicated to building an inclusive work environment. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity or expression, national origin, age, disability, status as a protected veteran, or any other characteristic protected by federal, state (including the California Fair Employment and Housing Act), or local laws.