Location: Remote Working
Start Date: ASAP
Duration: 12-Month Contract
Daily Rate: £450 - £600 per day outside IR35
Summary
Our client is building a Law Firm Digital Twin: a working simulation of how a law firm operates, built from the firm's own data. Data from the firm's operational systems (practice management and finance, HR, client onboarding, time recording) is mined into event logs; a discrete event simulation is built and calibrated from those logs; and the firm's financial statements are reconstructed on top, so the effect of an operational change can be quantified before it is made.
The stack spans three layers: data pipelines and a cloud warehouse (Snowflake), a simulation and statistical modelling core in Python, and a TypeScript product (Next.js, React, Postgres).
This role owns the data layer: the pipelines that turn raw system data into the event logs and modelling-ready structures everything else runs on. It is a hands-on building role in a small engineering team.
Key Responsibilities
- Design, build and run the data platform: the pipelines and warehouse that turn raw data from client systems into data the models and the product can use.
- Own data quality end to end, from source extraction through to the datasets the modelling team relies on.
- Make new data sources usable: understand what they hold, extract from them, and shape the results into the platform's structures.
- Work with the data scientist on what the models need, and with the software engineer on serving data to the product.
- Operate the infrastructure behind the data platform and keep it reliable, secure and affordable.
- Shape the data architecture as the platform scales.
- Above all, someone who has done analogous work: you have personally built the path from an enterprise's operational systems into a warehouse and into event-shaped analytical data. Specifically:
- A hands-on pipeline builder. You have built and operated production pipelines recently and can walk through one in depth: the source quirks, the failure modes, the validation, the re-runs.
- Warehouse and SQL depth. SQL as a first language, and real experience with Snowflake or an equivalent cloud warehouse: modelling, performance, cost, access control.
- Comfortable with messy enterprise extracts. Old operational databases, inconsistent fields, batch-stamped timestamps, undocumented conventions. You profile before you assume, and your validation catches what documentation misses.
- Careful with sensitive data. The pipelines carry confidential HR and financial data; access control and auditability are part of the engineering, not overhead.
- Python and/or TypeScript for pipeline code, with version control, tests and review as habits.
- AI-assisted engineering as a habit (Claude Code or similar), with judgement about where not to trust it.
- Experience of law firm or professional services systems (Aderant, Elite, Intapp, SuccessFactors, Workday)
- Process mining and event-log analytics (e.g., QPR, Celonis, Signavio)
- Process simulation.
