Fixed-scope automation for technical consultancies

AI and GIS workflow automation for technical consultancies.

I help engineering, environmental, energy and geospatial teams reduce repetitive data processing, technical-document review, mapping and reporting. Start with one fixed-scope working pilot delivered within ten business days.

Book a 30-minute workflow call View example automations

GISSatellite imageryTechnical documentsInternal AI tools

Fixed scope: clear deliverables
10 business days: working pilot
Pilot engagements: from €1,500

Kundan Thota

AI, GIS & workflow automation specialist

Ten-Day Automation Sprint

One workflow. Clear inputs. Clear outputs. Working pilot.

One repetitive technical workflow analysed and automated within ten business days.

Commercial terms

Fixed-scope pilot engagements from €1,500

50% at kickoff and 50% after the agreed acceptance test.

Suitable workflows

→Repeated GIS analysis and map production
→Satellite or spatial-data processing
→Technical PDF and document extraction
→Tender requirement and compliance extraction
→Automated project reporting
→Internal AI-assisted research tools

Deliverables

✓Workflow and bottleneck analysis
✓Working pilot
✓Documented inputs and outputs
✓Recorded handover
✓Fixed acceptance criteria

Problems I solve

Remove repetitive work from technical project delivery.

I support consultancies and technical teams delivering projects for municipalities, utilities, infrastructure providers, and public-sector clients.

Repeated GIS processing

Manual technical reporting

Tender-document review

Scattered project knowledge

Repeated data collection

Slow internal prototypes

Less manual work

Faster delivery

Reusable workflows

Traceable outputs

What a pilot looks like

Turn one manual process into a workflow your team can reuse.

Before

✓Files spread across folders
✓Repetitive manual processing
✓Outputs recreated for every project
✓No reliable workflow documentation

Pilot

✓Defined input
✓Automated processing
✓Review and quality-control step
✓Reusable workflow

After

✓Structured output
✓Repeatable execution
✓Reduced manual work
✓Documented handover

Example

Upload a tender package → extract mandatory requirements and deadlines → generate a traceable compliance matrix with source-page references and human review.

Example workflows

Automation first, backed by specialist geospatial and energy expertise.

Discuss a workflow

Workflow automation

Repeatable spatial delivery

GIS Workflow Automation

Automate repeated spatial joins, data cleaning, map generation, quality checks, and report outputs in a reusable workflow.

Data enrichmentQuality checksReporting

Document intelligence

Requirements you can trace

Tender Document Intelligence

Extract requirements, eligibility conditions, deadlines, and evidence into a traceable compliance matrix with source references and human review.

Document intelligenceTraceable outputsHuman verification

Rapid prototype

Working pilot in ten days

AI Prototype Sprint

Turn one clearly defined internal workflow into a working AI-assisted pilot with documented inputs, outputs, and acceptance criteria.

Internal toolsDecision supportHandover

Specialist service

Planning-ready heat layers

Municipal Heat-Planning Support

Estimate heat demand and enrich missing building attributes using GIS, satellite imagery, public datasets, and synthetic data.

Urban energyData enrichmentDecision support

Specialist service

Prioritised rooftop opportunities

Rooftop Solar Screening

Screen rooftops for solar suitability using building footprints, imagery, orientation, shading indicators, and automated prioritisation.

Spatial analysisOpportunity mappingAutomation

Specialist service

Usable data from scattered sources

Building & Urban-Energy Data

Gather, structure, validate, and enrich building-stock and energy data from fragmented public, spatial, and imagery sources.

Data acquisitionValidationData enrichment

Specialist background

Research depth with a practical delivery model.

I am an AI researcher at Karlsruhe Institute of Technology (KIT) and an independent specialist in GIS, satellite imagery, document intelligence, and technical workflow automation.

I work best with:

→Engineering, environmental, energy, geospatial, and infrastructure consultancies
→Teams delivering technical projects for municipalities, utilities, and public-sector clients
→Organisations that need a contained pilot before committing to a larger automation project

The outcome is not just a model or demonstration. It is a documented workflow, map, dataset, or decision-ready output that a technical team can evaluate and reuse.

Working arrangements

Based in GermanyRemote projects availableFixed-scope engagementsConfidentiality availableClient data is not reusedClear deliverables and acceptance criteriaInvoice-based B2B engagement

Background

✓Published research in geospatial AI and urban energy
✓M.Sc. Computer Science with applied AI focus
✓Experience building applied ML and automation workflows

Technical proof

Research that supports practical automation and analysis.

My published work focuses on satellite imagery, GIS, computer vision, synthetic data, and agentic data pipelines for city-scale building energy planning and data-scarce energy systems.

Urban heat demandVision-language modelsSatellite imageryGISBuilding energy planning

HeatPrompt: Zero-Shot Vision-Language Modeling of Urban Heat Demand from Satellite Images

Kundan Thota, Xuanhao Mu, Thorsten Schlachter, Veit Hagenmeyer

arXiv:2602.20066 [cs.CV]

Problem

Municipalities need accurate heat-demand maps for decarbonizing space heating, but many lack detailed building-level data.

Work

HeatPrompt uses pretrained vision-language models with an energy-planning prompt to extract semantic features from RGB satellite images, then combines those captions with GIS and building features for heat-demand estimation.

Result

The work shows that visual attributes from satellite imagery can improve heat-demand modeling in data-scarce regions and provide interpretable cues for city-scale building energy planning.

Commercial relevance

This workflow demonstrates how satellite imagery and building data can generate planning indicators without manually inspecting every location.

Read the paper

Building ageMulti-agent systemsComputer visionData fusionEnergy planning

A Multi-Agent System for Building-Age Cohort Mapping to Support Urban Energy Planning

Kundan Thota, Thorsten Schlachter, Veit Hagenmeyer

arXiv:2603.17626 [cs.CV]

Problem

Building-age distributions are crucial for heat demand mapping, retrofit prioritization, and district energy planning, but local datasets often contain gaps, inconsistencies, and fragmented sources.

Work

A multi-agent LLM system fuses Zensus, OpenStreetMap, and monument data, then uses BuildingAgeCNN with ConvNeXt, FPN, CoordConv, and SE blocks to classify building-age cohorts from satellite imagery.

Result

The pipeline creates structured building-age evidence, provides calibrated confidence estimates, and flags low-confidence predictions for manual review in planning workflows.

Commercial relevance

This approach can reduce manual classification and data-enrichment work when preparing city-scale building datasets.

Read the paper

Contact

Which technical workflow consumes too much of your team's time?

Send me the workflow, its inputs, current manual steps, and desired output. I will tell you whether it is suitable for a fixed-scope automation pilot.

Germany-based · Remote delivery · NDA available · Fixed-scope B2B engagements

Discuss a workflow