About ArborKnot

ArborKnot Capital Partners is an investment firm specializing in consumer debt portfolios across

Australian and European markets. Data is the core of how we operate: every portfolio we acquire is

priced, monitored, and optimized by models our data team builds and owns. We are a small, senior

team where your work reaches the investment committee, not a backlog.

The role

You will lead our data science function end to end. This is a hands-on leadership role: you own the

models that price every deal we do, the infrastructure that keeps them current, and the roadmap for

where our modelling capability goes next. We are building an AI-native data function: AI tooling is

load-bearing in how we work, and the next phase of our platform is an agentic layer that puts model

outputs directly in the hands of the business. You will work directly with the VP Data and the

commercial team, and mentor the data scientists working alongside you.

What you'll own

▪ Production models. Two models in production today drive portfolio pricing and recovery

forecasting. You own their accuracy, their maintenance, and their evolution.

▪ Retraining infrastructure. Build the pipeline that retrains our models on fresh data at a quarterly

cadence or better, so every pricing decision reflects current portfolio behavior.

▪ The feedback loop. Turn post-deal analysis into a systematic learning engine that sharpens

pricing transaction by transaction.

▪ The team. Mentor and grow the data scientists on the team; set the technical bar for how we

build.

▪ The agentic AI layer. Design and build the tooling that lets portfolio managers query data, run

scenarios, and get model outputs directly: agents, LLM workflows, and the governance that makes

them safe on real financial data.

What you bring

▪ 5+ years in data science with real production ML experience: models you built, deployed, and kept

alive

▪ Strong Python and SQL; comfortable owning pipelines, not just notebooks

▪ Experience with cloud data platforms (we run on GCP and BigQuery)

▪ Hands-on experience building with LLMs and modern AI tooling: you use AI daily in how you work,

and you've shipped real workflows or features on top of it

▪ Commercial instinct: you translate model outputs into decisions a non-technical stakeholder can

act on

▪ Ownership mentality: you close loops without being chased

▪ Fluent English (our leadership and counterparties are international)

Nice to have

▪ Experience designing agentic systems or LLM-based products running in production

▪ Background in credit risk, lending, collections, or debt purchasing

▪ Survival analysis, forecasting, or propensity modelling experience

▪ Experience leading or mentoring other data scientists