Python Developer

Brave; Cirencester; On-site; GBP 48,000 - 72,000; Full time

Benefits offered by this job

Pension Scheme

Healthcare Assistance Scheme

Birthday Off

Cycle to Work Scheme

Social Events

Additional 3 days holiday during shut/

Job summary

Landstack is a UK land and planning intelligence platform based in Cirencester. We are seeking a Python data engineer to own ingestion pipelines for planning data, enabling fast, reliable processing and delivery to customers.

You will build ETL pipelines, design data models in PostgreSQL/PostGIS, and optimize OpenSearch indexing. The role emphasizes data quality, monitoring, and thorough documentation within a collaborative engineering team.

Qualifications

• 3–5 years of professional Python experience, focused on data processing.
• Strong SQL and PostgreSQL experience; PostGIS a plus.
• Experience with ETL/pipeline patterns: batch jobs, queues.
• Familiar with pandas or polars and handling large datasets.
• Experience with AWS services and Docker/Kubernetes.
• Commitment to data quality and correctness.

Responsibilities

• Build and maintain Python ingestion pipelines for planning and geospatial data.
• Design data models and transformations for spatial data in PostgreSQL/PostGIS.
• Build and tune bulk indexing into OpenSearch.
• Add monitoring, validation and data-quality checks.
• Profile and optimise long-running batch jobs on AWS.
• Document sources, schemas and pipeline behaviour.

Skills

Python

SQL

ETL pipelines

Pandas/Polars

AWS & Docker/Kubernetes

Data quality

Tools

PostgreSQL

PostGIS

OpenSearch

Redis

SQS

Docker

Kubernetes

S3

Job description

Full Time: 37.5 hours a week Monday to Friday

• Pension Scheme
• Healthcare Assistance Scheme,
• Birthday Off,
• Cycle to Work Scheme,
• Social Events,
• Additional 3 days holiday during company shut down in December between Christmas and the New Year.

The Company:

Landstack is a UK land and planning intelligence platform used by developers, land agents, planners and housebuilders. We ingest and process planning applications, Land Registry title data, constraint layers, local plan policy and comparable transactions from hundreds of sources, and turn them into something people can search, analyse and act on. We're a profitable, bootstrapped team based in Cirencester.

The role

You'll own the pipelines that get data into Landstack: scraping and ingesting planning data from local authorities and national datasets, cleaning and normalising it, geocoding and spatially joining it, and loading it into PostgreSQL/PostGIS and OpenSearch. Much of this data is messy, inconsistent and large, and a lot of the value in the product comes from doing that well. You'll work directly with the founder and the rest of the engineering team, and you'll see your work land in front of customers quickly.

What you'll do

• Build and maintain Python ingestion pipelines for planning, title, policy and geospatial datasets
• Design data models and transformations for spatial data in PostgreSQL/PostGIS
• Build and tune bulk indexing into OpenSearch, including queue-based and S3 buffered ingestion paths
• Add monitoring, validation and data-quality checks so bad data is caught before it reaches users
• Profile and optimise long-running batch jobs for memory, throughput and cost on AWS
• Document sources, schemas and pipeline behaviour

What we're looking for

• 3–5 years of professional Python experience, with a real focus on data processing rather than general scripting
• Strong SQL and hands-on experience with PostgreSQL; PostGIS or other spatial experience is a big plus
• Experience with ETL/pipeline patterns: batch jobs, queues (Redis, SQS or similar), idempotent and resumable processing
• Comfortable with pandas/polars, pyarrow or similar, and with handling datasets that don't fit in memoryExperience scraping or integrating with awkward third-party sources and building resilience around them
• Working knowledge of AWS (S3, IAM, containers) and of running code in Docker/Kubernetes
• Care about correctness and data quality; you check the output, not just that the job finished