Mercury
Publicado originalmente en greenhouse · Aplica desde el sitio de la empresa para mejor experiencia.
<p>In 1965, an engineer in Scotland was given a mundane task with a tricky problem to it - banks wanted to close on Saturdays and still serve customers, but they didn’t know how to solve the authentication of the right user. James Goodfellow, working at Smiths Industries, uncovered the insights that you needed something you have (a card) and something you know (a PIN). Because someone told him they could only remember 4 digits instead of his proposed 6, today over 3 million ATMs use a card+4 digit PIN over 60 years later. </p> <p>Like the ATM, Mercury is building technology that pushes forward financial interfaces for the long term. Command is Mercury's LLM-powered financial assistant, launched to all customers in June 2026. It lets users understand their finances and take action in plain language, from asking about cash flow to sending payments, issuing cards, and managing invoices. With the product now in customers' hands, the work is to evolve it, extend its capabilities, and find new ways to leverage LLMs to give Mercury customers a more powerful banking* experience.</p> <p><strong>What you'll do</strong></p> <p>Ship new capabilities users love:</p> <ul> <li>Design and ship new Command skills, the domain-specific instruction sets that teach the model how to handle workflows like sending money, managing invoices, and understanding cash flow</li> <li>Design and build agentic workflows in Command, defining the architecture for how multi-step agent interactions should work as we extend what the product can do on a customer's behalf</li> <li>Work with backend teams to define tool schemas for new capabilities, shaping the data contracts between Mercury's business logic and the model</li> <li>Own new capabilities end to end, from the system prompt to the frontend component that renders the response</li> </ul> <p>Own the LLM layer:</p> <ul> <li>Maintain and evolve Command's prompt architecture: the system prompt, skill loading system, session context, and the policy and compliance layers underneath</li> <li>Tune model behavior: reasoning effort, prompt caching strategy, fallback chains, and the streaming patterns that make the product feel fast</li> <li>Stay current with how models are evolving and bring that knowledge back to how Command is built</li> </ul> <p>Build quality in:</p> <ul> <li>Write and expand Command's eval harness, adding cases that cover new capabilities and scoring rubrics that detect regressions before users do</li> <li>Partner with product and compliance teams to define what "working correctly" means for each new capability, then build the tests that prove it</li> <li>Own the reliability and quality of what you ship, from initial design through post-launch monitoring</li> </ul> <p>This list is illustrative. Command is a product in motion and priorities will shift as we learn. The right person will help choose the next highest-leverage work.</p> <p><strong>The ideal candidate</strong></p> <ul> <li>Has 7 or more years of software engineering experience, with deep technical expertise building and scaling LLM-powered applications in production</li> <li>Has gone beyond shipping a first version: you have scaled an LLM-powered product, dealt with the reliability and performance problems that come with real usage, and made it better over time</li> <li>Has experience designing agentic systems and has opinions about how to architect multi-step workflows that are reliable, explainable, and safe to run on behalf of real users</li> <li>Has built eval infrastructure and can write cases that actually measure whether the product works, not just whether the model outputs something plausible</li> <li>Understands the real tradeoffs in LLM deployments: latency, cost, compliance, and what breaks in production that doesn't show up in demos</li> <li>Has opinions about what makes an AI product trustworthy, not just impressive, and can build toward that bar</li> <li>Is comfortable with TypeScript and willing to learn Haskell for backend tool work, or already comfortable with both</li> <li>Can work across the full stack of an AI product, from the system prompt to the streaming frontend</li> <li>Has a track record of mentoring engineers and raising the technical bar of their team</li> </ul> <p>If this role interests you, we invite you to explore our public demo at <a href="http://demo.mercury.com">demo.mercury.com</a>.</p> <p>The total rewards package at Mercury includes base salary, equity, and benefits. Our salary and equity ranges are highly competitive within the SaaS and fintech industry and are updated regularly using the most reliable compensation survey data for our industry. New hire offers are made based on a candidate’s experience, expertise, geographic location, and internal pay equity relative to peers.</p> <p>Our target new hire base salary ranges for this role are the following:</p> <ul> <li>US employees (any location): $200,700 - $250,900</li> <li>Canadian employees (any location): CAD 189,700 - 237,100</li> </ul> <p>*Mercury is a fintech company, not an FDIC-insured bank. Banking services provided through Choice Financial Group and Column N.A., Members FDIC.</p> <p>Mercury values diversity & belonging and is proud to be an Equal Employment Opportunity employer. All individuals seeking employment at Mercury are considered without regard to race, color, religion, national origin, age, sex, marital status, ancestry, physical or mental disability, veteran status, gender identity, sexual orientation, or any other legally protected characteristic. We are committed to providing reasonable accommodations throughout the recruitment process for applicants with disabilities or special needs. If you need assistance, or an accommodation, please let your recruiter know once you are contacted about a role.</p> <p>#LI-RA1</p> <p> </p>