From Scattered Files to Scheduled Pipeline
Axis Spatial scans your existing workflow folder - ArcPy scripts, Excel files, QGIS projects - and reconstructs the actual process. AI agents interview your analysts to capture undocumented knowledge, then generate a unified Python pipeline ready to deploy on Databricks, AWS, or Azure.
SCAN
Static Analysis & Dependency Mapping
Point the tool at your workflow folder. The scanner identifies 30+ GIS file types, parses code to extract ArcPy operations, reads metadata, and maps how files connect to each other.
- Detects ArcPy, QGIS, FME, and Excel files
- Parses Python code using AST analysis
- Extracts shapefile schemas and CRS metadata
- Builds a dependency graph of your entire workflow
Runs locally. Your files never leave your machine.
$ axis scan ./workflows
Scanning directory...
Found: 12 .py files, 3 .qgz projects, 8 .xlsx
Parsing ArcPy operations...
47 files mapped, 23 connections
INTERVIEW
Logic Extraction & Knowledge Capture
AI generates questions based on what it could not figure out automatically. Why is the buffer 500 metres? Where does this data come from? Your analysts answer, capturing knowledge that exists only in their heads.
- Questions generated from code analysis gaps
- Multiple choice options to reduce typing
- Delegate questions to colleagues who know the answer
- Progress saves automatically - return anytime
Captures undocumented knowledge that survives staff turnover.
Interview Progress: 12/18
Q: Why is buffer set to 500m?
○ Regulatory requirement
● Species dispersal range
○ Legacy default
VALIDATE
Human-in-the-Loop Review
Before any code is written, you see the proposed workflow as a visual graph. Each processing step, each data dependency. Your team reviews, corrects mistakes, and approves the plan.
- Visual DAG shows all processing steps
- Edit nodes, add steps, change connections
- See how interview answers affected the plan
- Nothing happens without your sign-off
No black box. Full transparency before generation.
Workflow validated
├── Input: flood_zones.shp
├── Buffer → Clip → Reproject
└── Output: analysis.gpkg
Ready for code generation
GENERATE
Code Generation & Automated Testing
AI agents write Python code using open libraries - GeoPandas, Rasterio, GDAL. No ArcGIS license required to run. A second agent writes tests. Output is deployment-ready for your target platform.
- ArcPy translates to GeoPandas + Rasterio
- Output formats: GeoParquet, COG, STAC catalogues
- Automated tests validate correctness
- Terraform templates for infrastructure
You own the code. No vendor lock-in. Run on unlimited machines.
Generated: pipeline.py
import geopandas as gpd
from rasterio import open as rio_open
# No ArcGIS license required
Tests: 12 passing
DEPLOY
Cloud Infrastructure & Observability
Deploy to Databricks, AWS, GCP, Azure, or Snowflake - wherever your team already works. Includes scheduled job configuration, monitoring setup, and CI/CD templates.
- Databricks notebooks + Unity Catalog
- AWS Lambda + Step Functions
- Azure Functions + Data Factory
- Built-in monitoring and alerting
Integrates with your existing CI/CD pipeline.
$ axis deploy --target databricks
Packaging pipeline...
Uploading to workspace...
Schedule: daily 06:00 UTC
Deployment complete
Desktop ArcPy vs Automated Pipeline
| Aspect | Desktop ArcPy | Automated Pipeline |
|---|---|---|
| Concurrency | Single-threaded | Distributed (Spark) |
| Cost Model | Per-seat license | Compute-based |
| Environment | Local workstation | Cloud cluster |
| Monitoring | Manual checks | Built-in observability |
| Version Control | File timestamps | Git-based CI/CD |
Why GIS Teams Trust This Approach
Your Data Stays Local
File scanning runs entirely on your machine. Only workflow metadata is sent for AI analysis - never your actual geodata files.
Multi-Tool Understanding
Handles workflows spanning Excel, ArcGIS, QGIS, FME, and Python. Detects where data flows between tools and preserves that logic.
Knowledge Capture
The interview system extracts tacit knowledge: why parameters were chosen, what happens in edge cases. This survives staff turnover.
Tested Output
Every generated pipeline includes automated tests. Outputs are validated against expected results before production deployment.
Frequently Asked Questions
Does my code or data leave my infrastructure?
No. The scanner runs locally on your machine. Only workflow metadata (file names, detected operations) is sent to AI for analysis. Your actual geodata files never leave your environment. Generated code deploys to your own cloud account.
How long does the process take?
Scanning takes minutes. The interview phase depends on your team's availability - typically 1-3 days. Code generation takes hours. End-to-end, most workflows are automated within 1-2 weeks.
What if the generated code is wrong?
You review and approve the workflow plan before any code is generated. Generated code includes automated tests. You can request revisions or manually edit the output - you own the code.
Do I need ArcGIS licenses to run the output?
No. Generated pipelines use open Python libraries (GeoPandas, Rasterio, GDAL). Run on 100 machines without buying 100 ArcGIS licenses. Scale processing without per-seat costs.
What We Support
Input Formats
- ArcPy scripts (.py)
- QGIS projects (.qgz, .qgs)
- Excel with formulas (.xlsx, .xlsm)
- FME workbenches (.fmw)
- ModelBuilder exports
- Shapefiles, geodatabases, GeoPackage
Output Formats
- Cloud Optimised GeoTIFF (COG)
- GeoParquet
- STAC Catalogues
- PMTiles
- FlatGeobuf
- Delta Lake tables
Pick your messiest process. See results in days.
No 6-month proposal. No massive commitment.
NDA protection for all discussions.