Data Scientist

Data Scientist London Inside IR35 £700 per day 1 day per week on site I’m working with a London based team who need a Data Scientist to fix a forecasting model that’s failing in production. The foundations are there, but the pipeline isn’t stable and the business can’t rely on the outputs. They want someone who has walked into a messy modelling project before and turned it around. You’ll be owning the full diagnostics and rebuild of this model. What you’ll be doing • Pulling apart the existing forecasting and predictive models to understand where accuracy drops and why performance changes between training and live data • Rebuilding the modelling architecture in Python using best practice for reproducibility, versioning and testing • Working through feature engineering, identifying weak features, replacing low-quality inputs and strengthening the dataset • Running experiments to compare model versions, test different approaches and prove which solution gives the best long term stability • Deploying and monitoring the improved model in Azure so it runs reliably day to day • Working closely with engineering to streamline the end-to-end pipeline, from data ingestion through to model output • Putting proper monitoring in place so the team can spot drift, data quality issues or performance drops early Tech experience they want to see Python You need to be strong across modelling libraries, experimentation and writing clean code that the team can maintain. You’ll be rebuilding parts of the pipeline, not just tweaking notebooks. Azure You’ll be using Azure ML, pipelines, storage and deployment tools to get this model stable in production. If you’ve moved models from local development to Azure based production before, this is exactly that kind of work. Forecasting and predictive modelling This project needs someone who has solved real business forecasting problems. If you’ve fixed models suffering from drift, poor generalisation, weak features or inconsistent performance in live environments, you’ll fit straight in. This is for someone who enjoys stepping into the project when it’s at its messiest and getting it running cleanly. If you're an experienced Data scientist who feels like they can provide a solution feel free to apply or send your C.V to nokakpu@revoco-talent.co.uk Data Scientist