Geospatial Data Scientist

Geospatial Data Scientist

Who we are

At ArchAI, our mission is to uncover the lost landscapes that shape our present and future. We use AI to digitise Earth-observation data and historic maps at national scale, turning them into a growing suite of geospatial products that reveal lost woodland, ghost ponds, and archaeological traces for conservation and heritage teams, while giving planners and infrastructure partners the evidence to balance ecological opportunity with ground risk.

Your mission

As Geospatial Data Scientist, you will turn AI detections into dependable insight, moving seamlessly between deep-dive geostatistics, hands-on data clean-up, and concise storytelling. You will own the quality loop, shape the standards and workflows that keep projects on track, and collaborate with colleagues across modelling, product, and heritage to ensure every deliverable stands up to scrutiny and serves our diverse client base.

Responsibilities

• Load model outputs into QGIS, apply WMTS/WFS overlays, and run in-house Python QC scripts.
• Interrogate change statistically: run spatial-temporal queries to quantify change over time, connectivity shifts, fragmentation, and generate context-aware priority scores.
• Decide when to relabel data, brief subcontractors, or raise issues with the modelling team, documenting every step.
• Maintain a transparent priority list and flag shifts early, keeping delivery on track.
• Join discussions on how spatial evidence should guide policy or planning decisions, and present findings in plain English.
• Share interim results with academic, commercial, and charity partners, gathering their context to improve analysis.
• Tune the existing processing pipeline, evaluate new QGIS add-ons, and recommend visualisation upgrades that make findings faster to produce and easier to grasp.

Your profile

Must-haves

• Advanced QGIS user, confident with raster/vector processing and geostatistics.
• Practical experience with WMTS, WFS, and Python geoprocessing libraries (GeoPandas, rasterio, Shapely).
• Advanced spatial statistics—skilled in time-series analysis that quantifies connectivity and fragmentation, and produces rule-based priority metrics for change hotspots.
• Proven track record analysing large-area landscape or ecological change.
• Methodical, detail-oriented, and comfortable nudging the team when priorities drift.
• Clear written English and the ability to explain technical results to non-specialists.

Nice-to-haves

• Background in ecology, geography, wetlands, or heritage.
• Knowledge of UK rural landscape history (e.g., hedgerows, Cheshire & Norfolk ponds).
• Basic SQL/PostGIS.
• Experience translating spatial evidence into policy or planning recommendations.

What we offer

• Salary: £35 000 – £45 000 (dependent on experience)
• Permanent, full-time contract. Six-month probation
• Remote-first working with quarterly in-person team days in the UK.
• 28 days’ paid holiday a year, inclusive of public and bank holidays.
• Auto-enrolment pension with a 3 % employer contribution.
• A mission that matters: your analysis will steer national decisions on biodiversity, heritage, and ground-risk, giving you a direct line of sight from code to countryside.

Due to the high number of applications without answers to our questions, we will only consider candidates who email three separate documents to info@archai.io:

1️⃣ Your CV

2️⃣ A short cover letter

3️⃣ A single PDF answering any two of the questions below (≤ 300 words each)

1. Describe a time you spotted an unexpected pattern in spatial data. What tipped you off, and what did you do next?
2. Propose a simple rule-based system for ranking landscape-change hotspots for conservation urgency.
3. QC at scale: how would you design a workflow to validate AI-detected ponds across the UK?
4. Name the QGIS plugin or Python library that has most improved your large-scale spatial workflows and explain why.
5. A utility wants to plant 10 000 trees on historically wet land—outline your data-led case for wetland restoration instead.

No agencies, please.