Client-Facing AI Systems & Automation Engineer

Client-Facing AI Systems & Automation Engineer, Remote

We are Asterley Bros, a growing craft drinks business based in London. We make British vermouth, amaro, cocktails, and aperitivo products, selling through ecommerce, trade, events, distributors, and Amazon.

We are looking for a reliable, technically strong, client-facing engineer to help us build and support AI-enabled systems for both our own business and our Absolution Labs clients.

This is not a pure coding role. You will need to understand business problems, communicate clearly with non-technical people, train users, troubleshoot issues, and build systems that people actually use.

You should be comfortable moving between:

- building Python scripts, integrations, and AI workflows;
- joining client calls;
- listening carefully to what a client is really trying to achieve;
- explaining technical ideas in simple language;
- writing clear documentation;
- troubleshooting when something is not working;
- improving systems based on user feedback.

If you are technically strong but do not enjoy explaining, training, and supporting non-technical users, this is probably not the right role.

About us

Asterley Bros is a London-based craft drinks company. We make premium British vermouth, amaro, cocktails, and aperitivo products.


We are building practical AI-enabled operating systems, not demos. The goal is to create tools and workflows that save time, improve visibility, and help real businesses run better.

What you’ll do

You will work across internal Asterley Bros systems and client projects.

Your work may include:

- Mapping business workflows with our team or clients
- Building AI-enabled automations and integrations
- Creating internal tools, dashboards, approval flows, and reporting systems
- Integrating with tools such as Shopify, Google Workspace, spreadsheets, CRMs, Xero, email, and internal documents
- Building and maintaining Python scripts, APIs, and lightweight web tools
- Reviewing, testing, and improving AI-generated code
- Setting up logging, monitoring, and simple health checks
- Troubleshooting bugs, failed automations, broken integrations, and data issues
- Writing clear documentation and user guides
- Training non-technical users on how to use the systems
- Joining client calls to gather requirements, explain solutions, and support implementation

A typical project might be:

- Turning a manual spreadsheet process into an automated workflow
- Connecting Shopify, Google Sheets, email, and CRM data into one useful dashboard
- Building a tool that drafts customer or sales follow-ups for human approval
- Helping a client understand why an automation is failing and fixing it
- Training a client team to use a new AI-assisted workflow
- Reviewing AI-generated code, improving it, testing it, and deploying it safely
- Creating a simple dashboard or admin tool that non-technical users can actually understand

Client communication is a core part of the role

This is very important.

You must be able to communicate clearly with non-technical business users. Some of our clients will not know exactly what they need. You will need to listen carefully, ask good questions, and help translate their problems into practical technical solutions.

We are looking for someone who can:

- Listen well before jumping to a technical answer
- Ask clear, practical questions
- Explain technical issues in plain English
- Stay calm and helpful when troubleshooting problems
- Train clients patiently and clearly
- Write simple step-by-step instructions
- Tell us early when something is blocked, risky, or unclear
- Communicate progress without needing to be chased
- Be honest about what is possible, what is not, and what needs more investigation

If you prefer to only receive tickets and code alone, this is probably not the right role.

How we expect you to work with AI

We use AI tools heavily, and we expect this role to do the same.

This is not about using ChatGPT occasionally. We are looking for someone whose normal working cadence reflects modern AI-assisted software development.

You should be comfortable using AI coding tools such as Claude, ChatGPT, Cursor, GitHub Copilot, or similar every day to move faster on:

- boilerplate code
- API integrations
- tests
- documentation
- debugging
- refactoring
- code review
- implementation planning

We expect AI to increase your speed and output compared with a traditional hand-coding workflow. Modern AI tools remove a lot of slow boilerplate work, and we expect you to use them well.

But speed alone is not enough.

You are still responsible for the final result. You must understand the code, test it, review it, handle edge cases, protect secrets, and make sure the system is reliable and maintainable.

We do not want someone who blindly copies AI-generated code. We also do not want someone who works as if AI tools do not exist.

The right person will use AI as a serious productivity multiplier while applying strong engineering judgement.

What we expect technically

Required:

- Strong Python for scripting, APIs, automation, and backend workflows
- Experience consuming and building REST APIs
- Good understanding of authentication, API keys, OAuth, webhooks, error handling, and retries
- Comfortable working with JSON, CSV, spreadsheets, and structured data
- Experience with SQL or database-backed applications
- Git and GitHub workflow
- Basic cloud or server deployment experience
- Ability to test, debug, and maintain your own work
- Ability to write clear documentation for technical and non-technical users

Strong pluses:

- JavaScript or TypeScript
- Docker
- FastAPI, Flask, Django, or similar
- Google Workspace APIs, especially Sheets and Drive
- Shopify API
- Xero API
- CRM integrations
- Email integrations, IMAP/SMTP or Microsoft/Google APIs
- Firebase, Supabase, Postgres, or similar
- Simple web dashboards or admin tools
- Experience supporting clients or external users

AI and LLM experience

You should have hands-on experience using LLM APIs such as Claude, OpenAI, Gemini, or similar.

You should understand:

- Prompt design for real workflows
- Structured outputs, JSON schemas, and validation
- Tool calling or agent workflows
- Human approval gates
- How to constrain AI systems so they do not act unsafely
- How to evaluate and test AI outputs
- How to monitor failures and avoid silent errors
- Why AI-generated code still needs proper review, testing, and security checks

Experience with agent frameworks is useful, but not required if you are strong technically and understand the principles.

The mindset we want

This role needs someone who is:

- Reliable
- Detail-oriented
- A good listener
- A clear communicator
- Patient with non-technical users
- Comfortable with ambiguity
- Honest about what is working and what is not
- Good at asking questions
- Able to work independently
- Careful with sensitive business and client data
- Interested in building systems that real people use
- Willing to improve and maintain existing systems, not only build new things
- Fast-moving, but not reckless
- Comfortable using AI tools daily as part of their normal engineering workflow

We are not looking for someone who only wants to build exciting demos. We need someone who can help us create practical AI-enabled systems for real businesses.

Role details

- Location: Fully remote
- Company: Asterley Bros London, UK
- Type: Part-time or project-based to start, with potential to grow
- Hours: Flexible, but you must be able to overlap with UK working hours for part of the day
- Compensation: Competitive for the Philippines market, based on experience and ability
- Start: As soon as possible for the right person

How to apply

We receive a high volume of generic AI-written applications. Please read this section carefully.

Applications that look like generic ChatGPT cover letters, do not answer the questions below, or are not specific to this role will be ignored.

To apply, please send:

1. Your CV or resume
2. Your GitHub, portfolio, or examples of previous work
3. A short 3 to 5 minute Loom video, or similar screen-recorded video
4. Short written answers to the questions below

Applications without a video will not be reviewed.

Please start your application message with this exact sentence:

'I build systems people actually use'

In your video, please answer:

- Why are you interested in this role?
- Tell us about one automation, integration, or internal tool you have built. What problem did it solve, what tools or APIs did it use, and what was hard about it?
- Tell us about a time you had to explain a technical issue to a non-technical client, manager, or teammate.
- How do you use AI coding tools such as Claude, ChatGPT, Cursor, GitHub Copilot, or similar?
- Give one example where AI helped you build something faster than you could have by hand. What did you delegate to AI, what did you personally d ---------- , and how did you verify the result?
- How do you check whether AI-generated code is safe, reliable, and correct?

In your written answers, please answer briefly:

- What is your expected hourly or monthly rate?
- What hours can you regularly overlap with the UK working day?
- What are the main APIs, tools, or platforms you have integrated with before?
- If an automation fails in production, what are the first things you check?
- What parts of engineering should not be delegated blindly to AI?

Please do not send a long generic cover letter. A short, specific application is much better.

Generic applications will be ignored. We are looking for someone who has read the role properly and can show that they understand both the technical and client-facing parts of the job.