You and the process

With a person
Anna, in-house recruiter
9 years hiring engineers
30 minutes with a real recruiter
They read your CV with you, on a call, and say where the offers are being lost.
Didn’t find what you were looking for? Tell us what to build
Discovered Labs

Full Stack Engineer

  • Remote
  • 0-3 years

Salary

$20,000 - 28,000/ year

AI summary

For members

The whole posting in a few lines. Sign up to read it here and on every role you open.

Sign Up to Read

Description

The company helps $10M - $50M ARR companies get more leads, users and customers from Google, Bing, and AI assistants such as ChatGPT, Claude and Perplexity.

The company approaches marketing as an engineering system, with data in, insights out and feedback loops throughout. Its work covers data-driven automation, first-principles analysis of how search and AI systems behave, and ownership of the full funnel across several surfaces.

The company is building an agency OS, where the agency's work runs as software. Work can be handed to an agent once it can be verified, so each eval widens the set of tasks the team can trust an agent with.

Expert orchestrators supervise a fleet of agents running live client workflows. The system surfaces only the decisions that need a person, and each expert correction becomes an eval case.

You will join the Automations team, which builds the agent swarm: a growing set of agents and workflows that complete client work end to end, along with the engineering that makes them reliable. Your remit is to expand what the swarm can do and prove its output is correct, making non-deterministic systems reliable, measurable and provable.

You will work with the Automations lead and the AI & Data team, who own the platform and tooling. You take the work the last mile rather than building orchestration or infrastructure. The team is small and technical, with no layers of management, and you work directly with founders on architecture, code and product.

Accuracy of 80% means little if the agent cannot show which answers are the good ones. Accuracy of 95%, with visible reasoning and flagged uncertainty, is what makes an agent usable.

Responsibilities

  • Build eval harnesses and CI gates that catch agent regressions before release, including golden suites, deterministic checks and online scoring.
  • Add observability to agents: step traces, token and cost tracking, run-level SLOs and failure categories, so you can tell an agent that ran from one that was right.
  • Make every claim an agent produces traceable to its inputs, tool calls and reasoning steps.
  • Ship end-to-end agents with a review surface where experts approve, correct or reject work, and feed those corrections back as training signal.
  • Give the SEO team dashboards and controls to run, inspect and trust agent output without asking engineers.
  • Extend agents to open pull requests in real repositories, authenticate through OAuth, and operate real sites and CMSs through a browser.
  • Turn existing AI perception and citation data into client recommendations, prioritised actions and measurable results.
  • Implement scoring and ranking tools for internal linking, semantic relevance, Reddit opportunities, and content and citation quality. You build and productionise these tools but do not own the underlying research.
  • Build shared, reusable modules that agents and workflows both compose, so each improvement reaches every agent at once.
  • Take workflows that work for one client and make them run reliably for many, with per-client configuration, isolation and contained failures.

Requirements

  • 1+ years in software engineering, with recent hands-on time on production systems built on LLMs or machine learning.
  • Strong Python, React and TypeScript skills and solid systems fundamentals. You write production services, not notebooks.
  • Product taste: you would rather build one thing an expert opens daily than five that get clicked once.
  • Shipped LLM applications in production using agents, tool use, structured outputs, retrieval and prompt design, and kept them running.
  • Experience writing evals for non-deterministic systems and defining what correct means when many answers are valid.
  • Observability for AI systems with tools such as Langfuse, LangSmith, Braintrust or OpenTelemetry, including tracing, run inspection and cost and token accounting.
  • You can diagnose why an agent failed on one run in a hundred and turn that finding into a lasting check.
  • Pipeline orchestration with Airflow, Dagster, Temporal or similar, covering retries, idempotency, partial failure and resumption.
  • Third-party API integration that handles auth, rate limits, pagination and breaking changes.
  • You ship your own code: containers, CI/CD, deployment and credential handling, using the platform the AI & Data team provides.
  • Product sense for expert users. You have built tools domain experts use daily and know an unexplained output is unusable.
  • Collaborative in writing: you define contracts with infrastructure owners, document decisions, write clear specs and state tradeoffs.
  • A builder who ships working systems rather than planning endlessly, and an operator who keeps systems reliable after launch.
  • An owner who catches silent bad output, fixes it and builds the eval that stops it recurring.
  • Detail-focused: you would ship one provably right thing over ten probably fine ones.
  • Sceptical of your own output until it is measured, humble and curious, open to feedback, a first-principles thinker, and always improving.
  • Preferred: browser automation or authenticated web agent experience with Playwright, Puppeteer or computer-use models.
  • Preferred: applied ML in production for ranking, scoring, classification or recommendation.
  • Preferred: Python, React and TypeScript experience.
  • Preferred: prior experience at a fast-moving startup.

Conditions

Fully remote position.

You report to the CTO and work alongside the Automations lead.

Benefits

  • Work directly with the CTO on high-impact projects.
  • High ownership and autonomy, with no micromanagement.
  • First-hand exposure to cutting-edge AI and search technology.
  • Your work directly affects the performance of well-known companies with 10M+ ARR.
  • Join a fast-growing company at the intersection of AI and marketing.

About the company

Discovered Labs

Your chances

Worth a look before you spend an evening tailoring a CV for it.

  • 20 checks run
  • 3 red flags

Still hiring?

13 checks

1 red flag

How crowded?

7 checks

2 red flags

Fits Me

How well does this role fit you?

Answer a few questions or drop your CV, and every role gets a fit score with the reasons, this one first.

  • Your field
  • Level
  • Stack
  • Work model
  • Salary floor
  • Must-haves
Details8 facts · Role, Location, Compensation
Tech stack
  • React
  • TypeScript
  • Python
  • CI/CD
  • LLMs
Specialty
Fullstack
Region
Europe
Pay period
Annual
Show 4 more factsShow less

Role

Category
Development
Specialty
Fullstack
Experience
1+ years
Tech stack
  • React
  • TypeScript
  • Python
  • CI/CD
  • LLMs

Location

Work model
Remote
Region
Europe

Compensation

Salary
$20,000 - 28,000 / year
Pay period
Annual

Something wrong with this vacancy?

Similar vacancies

Share this vacancy

What's wrong with it?

The employer never sees who reported.

Reason