Founding Engineer

About Thunder

Thunder is building AI agents for B2B growth, starting with paid ads. Our platform gives growth marketers the same kind of leverage that engineers are getting from tools like Claude Code and Codex.

Rather than sell another piece of software into an already fragmented stack, Thunder sells the entire outcome. Customers give us responsibility for their paid channels, and we use our software to own strategy, execution, and results. This creates a tight feedback loop for improving our agents: whenever we encounter a recurring workflow, we look for a way to automate it and operationalize it across customers.

Thunder raised a pre-seed in late 2025 and is backed by early leaders at OpenAI and HubSpot. We're live with paying customers, from high-growth Series A-C companies to larger enterprises, and are using our product to run their paid channels end-to-end.

We're second-time founders. Between us, we've sold an AI startup, built core infrastructure for LinkedIn Ads, led product and growth for BuzzFeed Quizzes, and deployed AI agents at Fortune 500 companies.

About this role

We’re hiring a Founding Engineer to work directly with our CTO and founders on the systems that power Thunder’s growth marketing agents.

We’ve already built a durable agent runtime, persistent context and memory, remote execution environment, human-approval flows, and background routines. Our agents have more than 200 first-party tools for Google Ads and LinkedIn Ads alone, plus integrations across CRM, market research, search intelligence, and data analysis.

Growth marketing is a wide technical surface. It spans research and strategy, campaign creation, conversion tracking, analysis, reporting, experimentation, and ongoing optimization. Automating that work well requires agents that can reason over rich context, use many systems safely, run reliably over long periods, and involve people at the right moments.

Because Thunder runs paid growth for real customers, product development stays close to the work. You’ll see how operators solve problems today, ship new agent capabilities into live workflows, and quickly learn what works.

Specifically, you’ll:

• Build and improve the runtime for long-horizon, tool-using agents in sandboxed environments.
• Improve agent quality and reliability with SOTA continual learning, RL, fine-tuning, memory, context, evals, benchmarks, and production traces.
• Design novel interaction patterns and craft beautiful, resilient interfaces for human-agent collaboration.
• Give agents new capabilities through tools and integrations across advertising, CRM, analytics, research, and customer data.
• Ship end to end across infra, backend, and user-facing product experiences.
• Work with the founders and growth operators to turn real customer workflows into reusable product capabilities.

What we value

• Relentless Craft: We care about the details, from product design and interactions to system architecture and reliability. We study the history of software, prefer simple, elegant solutions, move quickly without creating unnecessary cruft, and leave every system better than we found it.
• Customer Obsession: We stay close to customers, understand what actually drives their business, and take responsibility for outcomes. When a problem repeats, we turn what we learned into better agent capabilities, tools, or workflows so customer work compounds into product.
• Deep Empathy: We're radically candid: we listen carefully, challenge each other's work directly, and assume positive intent. Feedback should make the work and team better; if we identify a problem, we help solve it, and mistakes become shared learning.
• Calm Intensity: We bring urgency, focus, and high standards to work. We move quickly without creating chaos. When things go wrong, we stay clear-headed, fix the problem, and learn from it.
• Founder Mindset: We take on ambiguous problems, define the scope, and take ownership for the outcome. We don't wait for perfect information or for someone else to unblock us; we move quickly from understanding the customer problem to shipping code.

What we’re looking for

We’re looking for a product-minded generalist who enjoys owning ambitious problems from idea through production. You should be deeply comfortable thinking through how an agent should behave, building the systems behind it, and shaping the experience around it.

Strong candidates will likely have many of the following traits:

• Experience owning and shipping meaningful production software end to end.
• Strong engineering fundamentals across APIs, data models, asynchronous systems, reliability, and debugging.
• Hands-on experience with LLMs or agents, especially tool use, orchestration, memory, context management, evals, or observability.
• Comfortable moving across the stack and learning unfamiliar systems when the work requires it.
• Strong product judgment and an instinct for turning ambiguous needs into simple, useful systems.
• Moves quickly while maintaining a high bar for correctness and security.
• Excited to work closely and in person with the founders and own important decisions from the beginning.
• Nice to have: previous founder, founding engineer, or early-stage startup experience.
• Nice to have: experience with TypeScript, Node, React, PostgreSQL, durable workflows, sandboxed execution, or complex third-party APIs.