
Saqara automates pricing file preparation for AOS
Fasfox designed an AI solution that reads incoming Excel price breakdown files, identifies their structure and converts them to the AOS format, reducing manual re-entry.
Saqara, a French software publisher specializing in the construction sector, is changing the way tenders are managed with its AOS platform (Appels d’Offres Simplifiés, or Simplified Tenders). The platform connects project owners with specialist construction companies across all trades, and makes it easier to select contractors for each work package, from concrete structures and electrical installations to finishing works.
Integrating DPGFs (décomposition du prix global forfaitaire, an itemized breakdown of a lump-sum price), sent by contracting clients as Excel files, previously required lengthy manual work to match the AOS platform format. Fasfox designed an AI model that handles this processing and frees the teams from repetitive data entry.
Missions
To automate the DPGF processing previously carried out manually by its teams, Saqara commissioned Fasfox to design an AI model able to process Excel files automatically. In time, the solution was to provide the following capabilities:
- Automatic analysis of the Excel files sent by contracting clients;
- Recognition and extraction of the relevant data from the specifications;
- Automatic formatting to the specific structure required by the AOS platform;
- Processing of the different types of work packages (structural works, façades, electrical, painting, doors, etc.);
- Standardization of work descriptions for consistent integration;
- Validation and quality control of the processed data;
- Automatic generation of formatted Excel files ready for integration into AOS.
Challenges

Saqara’s main challenge lay in the variety of Excel formats sent by contracting clients. Although all these files follow the same basic principle (listing line items that describe the supplies and works to be priced), their structure and organization varied considerably from one project owner to another.
This variety of formats created several major problems:
- Time-consuming manual processing: each Excel file required human intervention to identify its structure, extract the relevant data and reformat it to AOS standards.
- Lack of standardization: without a uniform format, the integration process could not be automated. Teams constantly had to adapt to each contracting client’s specific format, which slowed down the publication of tenders.
- Risk of human error: repetitive manual processing increased the risk of data-entry and interpretation errors, which could affect the quality of the data available on the AOS platform.
- Limited capacity to absorb growth: as the number of projects and contracting clients increased, manual processing was becoming a bottleneck.
The challenge was therefore to develop an AI solution able to handle any Excel file format intelligently, including unfamiliar ones, while ensuring reliable extraction and formatting of the data to AOS platform requirements.
Solutions

To meet these challenges, Fasfox developed an AI solution designed to automate Saqara’s DPGF processing, built around three areas:
- Development of an adaptive AI model: a system able to analyze and understand the structure of any Excel file automatically, so that it can adapt to each contracting client’s specific format.
- Process automation: the solution handles every step, from analyzing the input file to generating a file that meets AOS standards.
- Integration into the existing ecosystem: the AI model was developed to fit into Saqara’s existing ecosystem, so that it could be adopted without disrupting business processes.
Technical specifications
The solution developed by Fasfox is based on a technical architecture suited to processing heterogeneous Excel files:
AI model architecture
- Natural language processing (NLP) algorithms for the semantic analysis of work descriptions
- Pattern recognition techniques to identify the structure of Excel files automatically
- An automatic classification system for types of work packages and services
Data processing
- An Excel parser able to handle different types of data within the same file
- Algorithms to normalize and standardize the extracted data
- An automatic validation system to check the consistency of the processed data
Formatting and export
- A data transformation engine based on the predefined AOS format
- A generator of standardized Excel files with automatic formatting
Performance
- An architecture that allows several files to be processed in parallel
Results and benefits of the project

The deployment of the solution resolved Saqara’s initial problems with the manual processing of Excel files.
Results
Operational performance:
- Automation of the Excel file processing workflow: processing a file, which used to take several tens of minutes, now takes a few seconds.
- The model extracts and formats the data to AOS standards; an automatic validation step checks the consistency of the results before integration, which significantly reduces errors compared with manual data entry.
- Simultaneous processing of several files, whatever their original format.
- Delivery of an API integrated into an internal tool, so that staff can integrate DPGFs easily.
Adaptability and robustness:
- Handling of all DPGF Excel formats encountered to date, including previously unseen formats.
- Processing in production, regardless of the original structure of the files.
- Integration into the AOS ecosystem without disrupting existing processes.
Quality and reliability:
- Elimination of the data-entry and interpretation errors inherent in manual processing.
- Standardized output formats that meet AOS requirements.
- Full traceability of all processing carried out.
Benefits
For Saqara, automation frees the teams from repetitive data entry and makes it possible to handle more files without increasing headcount at the same rate. The costs of manual processing fall accordingly.
For contracting clients, tenders are published more quickly and with more predictable processing times. For construction companies responding to tenders, access to new opportunities is faster.