Applied AI Engineer - Direct Acquisition

Applied AI Engineer - Direct Acquisition
We're planning this as a six-month secondment. There may be an opportunity to extend it or make it permanent.

We're looking for an Applied AI Engineer to join the Direct Acquisition team.
You'll design and build AI agents that run whole business processes, from planning the work to learning from the results. What you find will steer how far iwoca takes agents across the business.

The company
Small businesses move fast. Opportunities often don't wait, and cash flow pressures can appear overnight. To keep going, and growing, SMEs need finance that's as flexible and responsive as they are.
That's why we built iwoca. Our smart technology, data science and five-star customer service ensures business owners can act with the speed, confidence and control they need, exactly when it's needed.
We've already cleared the way for 100,000 businesses with more than £5 billion in funding. Our passionate team is driven to help even more SMEs succeed, through access to better finance and other services that make running a business easier. Our ultimate mission is to support one million SMEs in their defining moments, creating lasting impact for the communities and economies they drive.

The team
Direct Acquisition includes iwoca's Marketing, Sales, Brand and Acquisition Tech teams. Together, they reach UK businesses of all shapes and sizes through paid and organic marketing channels and direct sales.
A new team in Direct Acquisition builds agents that automate and improve acquisition processes. That's its only focus, and it works closely with Acquisition Tech. It's small – a product owner and you – and everyone builds. You'll report to the Head of Direct Acquisition, who sets the commercial direction.

The role
As the Applied AI Engineer, you'll build the agents and the systems around them. Most of the work is system design and backend engineering.
You'll start with direct mail. The first goal is an agent that plans each monthly round, uses iwoca's existing predictive models to build the mailing file, and learns from the results as they come in. The pattern is simple to say and hard to get right: the agent orchestrates, deterministic code does the work, and a human in the loop approves the file before it goes to print. You'll work with the people who run each process to understand the edge cases, and with Core Systems and Compliance on access and controls.
The agents don't exist yet, and much is still open: which processes the team takes on, and what the architecture looks like. You'll own those design decisions and write most of the code yourself, much of it with AI coding tools. Some of this won't work, and finding out why is one of the main outputs of the role.

• Architecture. You'll design what agents need to run reliably: memory that persists between runs, task management, access to tools, permissions, and a record of every decision and why. You'll build each piece so the next process can reuse it.

• Guardrails. You'll decide where the agent uses judgement and where code enforces rules the agent can't change, like suppressions, compliance checks and budget limits.

• Delivery. You'll own each agent end to end: connect it to iwoca's predictive models, data pipelines and channel tools, and run it alongside the current process until it has shown it works.

• Measurement. You'll track how often agents need a person, where they fail, what they cost, and what they're worth.

The requirements
We're looking for the depth of someone who has solved hard problems, and the pace of someone who moves fast in areas that are new to them.
Essential:

• System ownership. You've owned a complex system end to end, from design to running it in production, and lived with the consequences of your architecture choices.

• Design judgement. You've made architecture decisions other people built on, and you can explain the trade-offs and what you'd change. When an idea won't work, you say so and offer a better one.

• Pace. You've got things live quickly in unfamiliar areas and they kept working once people relied on them.

• Agent engineering. You've built systems where LLMs take actions – more than chat or summaries – and you can explain what went wrong and how you fixed it. You know when a problem needs an agent and when it needs a for-loop.

• Controls. You've built permissions, data access rules and audit trails into systems from the start.

• Constructive challenge. You work well with commercial and product colleagues, and when an idea won't work, you say so and offer a better one.

• AI leverage. You've used AI coding tools to produce engineering work you couldn't have delivered without them – a service, an integration, or an internal tool others use. You have opinions on how to get the most out of them, and you check their output before it reaches production.

• Communication. You write and speak clearly, directly, and concisely. You adapt technical detail to your audience.

Bonus:

• Languages and data tools. You've worked with Python, Go, Snowflake or dbt.

• Evaluation. You've built evals that measure how an agent performs and where it fails.

• Marketing and growth tech. You've worked with ad platforms, CRMs or direct mail.

• Quantitative foundations. You come from a numerical or technical field, such as physics, engineering, maths or computer science.…

Source: Arbeitnow (https://www.arbeitnow.co.uk/jobs/companies/iwoca/applied-ai-engineer-direct-acquisition-london-36548)