Lead AI Engineer

# **Lead AI Engineer** ## **About Sierra Studio** When you apply to Sierra, you join an ecosystem. We specialize in connecting talented Brazilian professionals with exciting career opportunities in a highly-vetted small community of growing companies in the US. Applying to Sierra seamlessly integrates you into this ecosystem, granting automatic eligibility for all relevant opportunities we offer. **Location:** Brazil, remote **Compensation:** $70k – $120k USD, CNPJ **Reports to:** Founder & CEO ## **About our hiring partner** Our hiring partner is a US-based retention marketing agency helping DTC brands build personal relationships with their customers at scale through the channels they own: email, SMS, and everything in between. Their thesis is simple: brands shouldn't have to grow by unprofitably spending more on paid social. Everything they do is built around producing programs and messages that customers actually want and that actually perform. They're a team of about 12 (strategists, specialists, designers, copywriters, and ops), fully remote, at ~$1.2M in revenue with a goal to double in the next four months. They partner with the leading platforms in the space and are now evolving from an agency with great internal tooling into something closer to an agency and a software company combined. ## **About the Role** Our hiring partner is hiring their first engineer. The founder has already built the foundation himself using AI coding tools: a data layer that pulls Klaviyo data for every client into the company's own Supabase database (so the team can query it in ways the platform's native API doesn't allow), a set of AI skills and workflows the team uses instead of building reports by hand, and a diagnostics tool made of agents that continuously QA client accounts and flag what's broken. On top of that sit early products that are somewhere between prototype and solid foundation. It works and it's validated. But it was built by one person, and it's reached a level of disorganization that isn't sustainable as the company scales. The products need to be rebuilt with scale in mind, and the whole system needs an engineer who owns it. That's where you come in. You'll own the architecture end to end: auditing what exists, deciding what stays and what gets rebuilt, and turning it into a fully maintained codebase with clear guidelines. You'll pair closely with the founder, who leads on product while you lead on development, iterating on agency workflows and products with feedback that comes directly from clients. And you'll build agentic flows across nearly every area of the business, from account diagnostics to go-to-market, which today are all built and maintained by the founder alone. This is a lead role, not a founding-engineer-in-name-only role: you set the technical direction and standards for everything that gets built, and as the engineering function grows, you'll be the one shaping it. ## **What You'll Own** ### ***Architecture and Codebase*** - Audit everything that's been built so far, define what stays, what gets rebuilt, and how it all fits together. - Take the current system (hosted on Cloudflare, with Supabase as the data layer, all connected) to a fully maintained, documented codebase with clear standards and guidelines. - Design the data pipelines: ingestion from Klaviyo, Postscript, and Attentive, modeling and storage in Supabase, syncing, and querying beyond what the platforms' APIs allow. - Build for scale and clarity, so the business has a foundation it can actually execute on rather than a pile of scripts. ### ***Agentic Workflows and Internal Tooling*** - Design and ship agents and AI skills for account diagnostics, reporting, QA, go-to-market, and internal ops. - Make internal AI tooling reliable and consistent, so the ops team can roll it out across the team with confidence. - Define the quality boundaries so the whole team gets the same results the founder gets. ### ***Products*** - Evolve the early products built on top of the data layer into something clients can depend on. - Own the integrations with the email and SMS platforms the agency works with, and whatever comes next. - Establish the workflow between you and the founder: what he ships himself with AI coding tools, what goes through you, and how changes make it into the maintained codebase. ### ***Being the Technical Voice in the Room*** - Be the engineering counterpart to the founder in product decisions, giving honest reads on feasibility, effort, and risk. - Get in front of clients when it matters, in a product-manager-style loop: understand the problem, adjust, ship, repeat. - Set the engineering standards and practices that future hires will follow. ## **Qualifications** ### ***Required*** - 5+ years building software professionally, with a track record of shipping things real people used. - Strong full-stack engineering skills (TypeScript and/or Python, modern backend and frontend, Cloudflare, Supabase/Postgres, SQL). You write code every day and you're good at it. - Cloud infrastructure provisioning experience (AWS, GCP, Azure,..). - CI/CD workflows experience. - Hands-on experience building LLM-based systems: agents, tool use, RAG, evals, prompt and skill design. You understand where these systems break, not just how to call the API. - Good taste for data pipelines: pulling data from third-party APIs, modeling it, keeping it in sync, and querying it well. - Comfortable with Supabase/Postgres (or equivalent) and owning data infrastructure end to end. - Experience being the only or lead engineer somewhere: you can scope, prioritize, ship, and maintain without a team around you. - Ownership mindset. You say what you're going to do, and you do it. No one is going to check whether you're doing enough work. - Proactive communicator in a remote, async environment, with excellent written and verbal English. You'll work directly with the founder and, over time, with clients. ### ***Preferred*** - Experience with email/SMS marketing platforms (Klaviyo, Postscript, Attentive, Braze) or the DTC/ecommerce ecosystem. - Background at a fast-moving martech or retention product company. - Experience productizing internal tooling, or taking an agency's workflows and turning them into software. - AI coding tools (Codex, Claude Code, or similar) are already part of how you work. - Product instinct: you have opinions about what should be built, not only how. ## **What Success Looks Like** - The architecture has been audited, cleaned up, and documented, and the codebase is fully maintained with clear guidelines. - The internal AI tooling is reliable enough that the team gets the same quality of results the founder gets, without needing him in the loop. - Agentic workflows across the business are built and maintained by engineering, not by the founder on the side. - The early products have moved from prototype to something clients rely on, built to scale. - The founder spends his time on product, clients, and growth, because engineering is covered. - You've laid the groundwork for the engineering function to grow beyond one person. ## **Benefits** - 23 days of PTO. - Generous maternity and paternity leave. - Bonus structure. - Fully remote with flexible hours and US overlap. - Contractor engagement (no healthcare coverage). ## **A Note on Fit** The culture is built around ownership. Everyone geeks out on their craft, shares what they learn, and treats each other like adults. Because the team is remote, there's a lot of flexibility and an equal amount of accountability. People who take a full area, own it end to end, hit deadlines, and communicate proactively do very well here. People who need to be monitored don't. What makes this role different is where you sit. Most engineers building in this space are inside a software company, one step removed from the people using what they ship. Here you're inside the agency: the feedback loop with clients is direct, the results are measurable in revenue, and you're never limited to what the software can do. You get to own an entire engineering function and be at the front of figuring out what it looks like when an agency and a software company become the same thing. If that sounds like the kind of problem you'd want to go build your way through, you'll like it here.