Estimate AI
Construction, Estimating
An extraction platform that trains a dedicated model for each vendor, turning inconsistent construction estimate PDFs into structured, ERP-ready data.
Web Applications & Integrations
Field-to-finance ETL for a construction company
An Apache Airflow pipeline that moves the hours, machine usage and materials logged in SmartDok into Duett for payroll, timesheets and invoicing, with no manual re-entry.
- Field data reaches payroll without manual re-entry
- Automated
- Unmapped or invalid records are held back and logged
- Validated
- Every sync logged with its state and ID mappings
- Auditable
The starting point
The challenge
Field workers logged project hours, machine usage and materials in SmartDok, while finance ran payroll and invoicing in Duett. Exporting and re-entering the data by hand caused delays, bottlenecks and payroll errors.
The solution
What we built
Extracted project hours, machine usage and materials from the SmartDok REST API, in both JSON and XML.
Normalised the hierarchical payloads, and mapped employee IDs, cost codes and project identifiers between the two systems.
Validated every record against field rules and the ID mappings, isolating discrepancies in error logs instead of passing them to payroll.
Loaded verified records through the Duett REST API to update payroll, timesheets and invoicing.
Orchestrated scheduled syncs and retries with Apache Airflow DAGs, with PostgreSQL storing sync state, ID mappings and audit logs.
Architecture
How the system works
Orchestrated by Apache Airflow, with sync state and audit logs in PostgreSQL, running in Docker on AWS.
Engineering
How it was engineered
Kept invalid or unmapped records out of payroll: they are isolated and logged for review.
Logged every sync in PostgreSQL, so each record can be traced from SmartDok to Duett.
Containerised the pipeline with Docker and deployed it on AWS.
Scope
What was delivered
01
SmartDok extraction
02
Parsing and identifier mapping
03
Validation and anomaly checks
04
Duett loading
05
Airflow orchestration and audit logging
Technology
Stack
PythonApache AirflowREST APIsJSON/XML parsingPostgreSQLDockerAWS
More work
Related case studies
Enterprise integration hub
Enterprise integration
A secure, reliable integration layer connecting Salesforce with the enterprise's other platforms through MuleSoft, REST APIs and Apex callouts. It is designed for dependable data exchange, clear data ownership, and error handling that keeps operations running.
Armor Wallet
Crypto, DeFi, Web3 infrastructure
An open-source MCP server and agent layer that lets AI assistants carry out real crypto operations: wallet management, swaps, DCA and limit orders, and staking.
Need something similar?
We can scope and build a system with comparable capabilities, shaped around your own workflows, data and integrations.