SYSTEMS · DIGITAL TWINS · GIS · SIMULATION · REAL-TIME DATA · DEVOPS

SOFTWARE ARCHITECT · TECHNICAL LEAD · LEAD BACKEND ENGINEER

I build software for the systems society depends on.

Digital twins, scientific simulation, geospatial intelligence and real-time platforms for critical infrastructure, public services and crisis response.

From technical direction and hands-on backend development to integration, DevOps and continuous operation.

01 Architecture02 Geospatial03 Simulation04 Real-Time Data05 Integration06 DevOps
National-scale systemsTransport, health, water, energy and government
Critical infrastructurePower, water, transport, telecoms and emergency response
Real-time & scientificIoT, simulation, streaming and analytics
End-to-end ownershipArchitecture, backend, integration and DevOps

FEATURED WORK

Operational systems, not isolated demos.

Selected case studies showing architecture ownership, backend depth, scientific and geospatial engineering, cross-system integration and production delivery.

Browse the complete project library 10 flagship stories · 39 structured project entries

CHATBOTS & CONVERSATIONAL SYSTEMS

Conversational software for real guest-service workflows.

I designed and implemented a hotel chatbot platform that maintains multi-turn state, routes intents and entities, connects to hospitality systems, supports real-time communication and hands conversations to hotel staff when automation is not enough.

HOTEL VIRTUAL ASSISTANTguest-service workflow
Can I book a table for four tonight and request a late checkout?
I can help with both. I’ll check restaurant availability first, then continue with your stay request.
route restaurant_reservation · party_size=4
context retain_guest_session · continue_stay_service
Your restaurant request is ready. For late checkout, I can connect you with reception if the availability check cannot confirm it.
Conversation → hotel workflow → staff handoff

Implemented capability

The backend connects conversation flows with reservations and guest services through APIs, asynchronous messaging, persistent session state and PMS/POS integrations.

Chatbot architectureMulti-turn contextIntent & entity routingSession statePMS / POSREST & WebSocketsAsync messagingOmnichannelMultilingualHuman handoff
View the verified chatbot project
CHANNELweb / messaging
SESSIONcontext + state
ROUTINGintent + entities
HOTEL SYSTEMSPMS / POS / booking
RESPONSEautomation / handoff

CAPABILITIES

What I can lead.

The domains change. The recurring work is turning difficult, data-heavy and operationally sensitive problems into systems that can be deployed, extended and trusted.

01

Architecture & Technical Leadership

Modular systems, explicit boundaries, data ownership, security models and deployment strategy.

02

Backend Platforms & APIs

Python, Django, DRF and FastAPI platforms with complex domain logic and asynchronous workflows.

03

Geospatial Intelligence

PostGIS, raster and vector processing, map services, spatial filtering and real-time GIS.

04

Digital Twins & Simulation

Real infrastructure data connected to scientific models, scenarios, visualization and decisions.

05

Scientific Computing

Numerical processing, optimization and reproducible simulation workflows moved into production systems.

06

Real-Time Data & IoT

Telemetry, smart meters, public transport and sensor ingestion with time-series analytics.

07

Critical Infrastructure

Operational and decision-support systems for power, water, transport, telecoms and public safety.

08

Crisis Response

Rapid, reliable systems for health emergencies, biosecurity, wildfire and disaster operations.

09

Conversational Systems

Multi-turn chatbot backends, hospitality integrations, real-time messaging and human handoff.

10

DevOps & Infrastructure

Docker, Linux, reverse proxies, CI/CD, monitoring, backups and multi-environment operations.

ARCHITECTURE UNDER PRESSURE

The diagram is the easy part. The system has to survive reality.

I design around the conditions that expose weak systems: fragmented inputs, expensive computation, spatial complexity, organizational handoffs, changing rules and production failure.

01
REALITYFragmented sourcesFiles, sensors, external APIs and inconsistent records
ARCHITECTURAL RESPONSEEstablish an authoritative truthValidate, normalize, queue and preserve provenance
OPERATIONAL RESULTTrusted servicesConsistent APIs, analytics and downstream decisions
02
REALITYExpensive computationScientific models, optimization and long-running jobs
ARCHITECTURAL RESPONSESeparate interaction from executionQueues, workers, reproducible inputs and result stores
OPERATIONAL RESULTResponsive systemsTraceable runs without blocking users or services
03
REALITYSpatial complexityNetworks, rasters, coordinate systems and live positions
ARCHITECTURAL RESPONSEMake location first-classPostGIS, spatial services, transformation and caching
OPERATIONAL RESULTLocation-aware actionMaps and workflows that reflect the physical world
04
REALITYOrganizational handoffsDifferent authorities, roles, rules and responsibilities
ARCHITECTURAL RESPONSEMake boundaries explicitContracts, permissions, states and accountable transitions
OPERATIONAL RESULTVisible ownershipEvery action has a responsible role and history
05
REALITYRules keep changingPolicies, scenarios and operational needs evolve
ARCHITECTURAL RESPONSELocalize the changeModular domain logic, state machines and clear interfaces
OPERATIONAL RESULTAffordable evolutionNew requirements without destabilizing the platform
06
REALITYProduction will failInfrastructure, integrations and dependencies degrade
ARCHITECTURAL RESPONSEDesign for recoveryContainers, observability, backups and repeatable delivery
OPERATIONAL RESULTRecoverable operationsFailure is visible, bounded and rehearsable
THE STANDARD

Failure should be visible. Change should be affordable. Recovery should be rehearsed.

TOOLS FOLLOW THE PROBLEM

BackendPython · Django · DRF · FastAPI · Celery
DataPostgreSQL · PostGIS · TimescaleDB · Redis · InfluxDB
GeospatialGDAL · Rasterio · GeoPandas · GeoServer · Leaflet · CesiumJS
IntegrationKafka · MQTT · WebSockets · REST · SOAP · GTFS-RT · SIRI
ScientificNumPy · SciPy · Pandas · optimization · simulation wrappers
InfrastructureDocker · Linux · Nginx · HAProxy · CI/CD · monitoring · backups

ENGINEERING UNDER PRESSURE

Some systems become operational while the crisis is already happening.

01SNOWNational COVID case surveillance
02ERLYPopulation-scale diagnostic-test ingestion
03TCDC / EDVCSecure European certificate issuance workflows
04FMDEmergency animal-movement control
05HephaestusPredictive wildfire simulation

These projects combined rapid architecture, secure data handling, reliable deployment, operational feedback and continuous adaptation as rules and conditions changed.

RESEARCH → OPERATIONS

Models become valuable when they survive real users, data and infrastructure.

HephaestusWildfire scienceOperational simulation engine
CAT-BRAINCognitive rehabilitation researchClinical SaaS
EAC UFLSPower-system stability researchProtection decision support
SoteriaRoad-safety analyticsEngineering prioritization
CyDT / CITWINDigital-twin researchScenario and resilience platforms
AutoTRUSTConnected-mobility researchIn-vehicle IoT interfaces

ABOUT

Architecture is only useful when it survives contact with reality.

I am a software architect and technical lead focused on complex operational software. My work combines backend systems, geospatial technologies, scientific models, real-time data, distributed infrastructure and decision-support workflows.

I have worked as Software Architect, Lead Backend Developer, Technical Lead and DevOps Lead, often owning the path from early system design through production deployment and long-term evolution.

Where I add the most value

  1. 01
    Turn ambiguity into architectureModel domain requirements, ownership, interfaces and failure modes.
  2. 02
    Build the difficult backend pathData models, APIs, asynchronous pipelines and complex domain logic.
  3. 03
    Connect models and infrastructureGIS, telemetry, simulators and external enterprise systems.
  4. 04
    Move research toward operationsHarden prototypes for real users, data and deployment constraints.
  5. 05
    Own production deliveryContainers, CI/CD, reverse proxies, monitoring, backups and troubleshooting.

COLLABORATE

Building something technically difficult?

I’m interested in platforms involving infrastructure, simulation, spatial data, real-time telemetry, complex integrations, operational workflows or conversational interfaces.

Explore the workProfessional collaboration · Research partnerships · Consulting