Company Description
We are a digital growth company running two sides of one business: a performance marketing practice serving clients, and our own product built in-house. Both rest on the same belief — that most of the work people grind through by hand today should be handled by systems that run on their own.
Ideas move quickly from concept to production. What you build goes live, gets used daily, and gets judged on results.
Role Description
You will design, build, and own LLM-powered agents and automations that run in production and carry real operational load.
The job is to find where the business is slow, build the thing that fixes it, connect it to the platforms and services we already run on, and keep it working reliably once people depend on it. You will also help maintain the backend services behind those automations.
Building is not the whole role. AI tooling and search behaviour shift week to week, and we want someone who follows it closely, tests it personally, forms an opinion, and brings us proposals — ideally with a working prototype already attached.
----------------------------------
Key Responsibilities
Build and own agents in production
- Own our production agents and chatbots end to end — the prompts behind them, the tools and functions they call, and how their conversations flow
- Take a messy business problem and turn it into an agent that solves it: break the task down, connect the tools, APIs and knowledge sources it needs, and orchestrate the steps
- Test properly — run prompt experiments quickly, build evals so quality is measured rather than assumed, and ship improvements on evidence
- Keep what is live healthy: watch for failures, debug them fast, and keep raising reliability and guardrails as usage grows
Automate real work
- Put agents inside the workflows people already follow, so manual steps disappear rather than get a nicer interface
- Own the outcome, not the demo — hours saved, turnaround time, error rate, adoption
- Work with the people doing the work to decide what is worth automating, build rough prototypes fast to test the idea, and turn the ones that prove out into tools the team uses daily
Build the plumbing
- Build and maintain the integrations these automations depend on: third-party APIs, CMS and platform APIs, internal services, webhooks, and data pipelines
- Handle the unglamorous side of integration work — rate limits, pagination, auth and token refresh, retries, schema changes, and data that arrives dirty
- Handle day-to-day backend work on our product: API endpoints, database queries and migrations, background jobs, scheduled tasks, logging and basic monitoring
Stay ahead
- Track what is genuinely new and what actually matters — new models and their pricing, agent tooling, and how search and AI-driven discovery are changing
- Test new tools yourself instead of trusting the announcement, and form a view
- Bring proposals to us unprompted, with a prototype or a clear cost-and-benefit case behind them
----------------------------------
Qualifications
- Bachelor's degree or higher in Computer Science, Engineering, Data Science, or a related field — or equivalent demonstrated ability. What you have built matters more to us than where you studied
- Hands-on experience building with LLM APIs: prompt engineering, tool and function-calling, RAG, or agent frameworks
- Strong Python, and confident working with APIs to build and ship integrations end to end
- Working backend ability — a backend framework, SQL, Git, and basic deployment; comfortable moving through a codebase you did not write
- Practical experience integrating with third-party APIs, and with at least one CMS or platform API
- A track record of shipping — production work, client work, hackathon wins, or personal builds. Send GitHub, live demos, or write-ups
- AI-native: fluent with modern AI tooling and fast at turning an idea into something that runs
- Genuinely current — you can tell us what changed in AI or search in the last month, and why it matters
- Bias for action, comfort with ambiguity, and clear communication with non-technical colleagues
----------------------------------
Good to have
Backend & Engineering Depth
- Backend architecture and problem-solving — able to reason about how services fit together, weigh trade-offs, and debug problems you have not seen before
- Go & TypeScript
- PostgreSQL
Reliability, Security & Cloud
- Production reliability and security — monitoring and alerting, incident handling, secrets management, access control, and safe handling of user data
- Cloud and deployment experience — containers, CI/CD, and running services on a major cloud provider
- Alibaba Cloud experience specifically
- Digital marketing or SEO exposure
Domain & Tooling
- Digital marketing or SEO exposure
- Eval frameworks and structured LLM testing
- Scraping and crawling at scale
- Enough frontend ability to build a usable internal tool