Performance impact of optimizations
Challange
Growing system from ERP segment with many time-consuming computations blocking users from doing their job ⚡
Growing system from ERP segment with many time-consuming computations blocking users from doing their job ⚡
We performed several iterations of algorithm optimization, producing actual effect already on 2nd iteration - all in less than a month
| CHUNK SIZE | MEAN (MS) | ERROR (MS) | STDDEV (MS) | ALLOCATED (MB) |
|---|---|---|---|---|
| 50,00 | 1,792.7 | 19.69 | 18.42 | 254.72 |
| 100,00 | 1,804.8 | 13.18 | 11.680 | 254.75 |
| 250,00 | 1,813.6 | 7.680 | 6.820 | 254.73 |
| 500,00 | 1,791.8 | 18.17 | 15.17 | 254.74 |
| CHUNK SIZE | MEAN (MS) | ERROR (MS) | STDDEV (MS) | ALLOCATED (MB) |
|---|---|---|---|---|
| 50,00 | 2,076.6 | 36.46 | 39.01 | 39.01 |
| 100,00 | 2,007.9 | 19.14 | 15.98 | 15.98 |
| 250,00 | 1,965.3 | 18.77 | 16.64 | 16.64 |
| 500,00 | 2,176.3 | 39.16 | 76.38 | 76.38 |
| CHUNK SIZE | MEAN (MS) | ERROR (MS) | STDDEV (MS) | ALLOCATED (MB) |
|---|---|---|---|---|
| 50,00 | 902.8 | 17.80 | 34.29 | 268.69 |
| 100,00 | 861.0 | 16.78 | 28.040 | 261.44 |
| 250,00 | 919.1 | 18.090 | 22.22 | 257.32 |
| 500,00 | 1,297.8 | 20.20 | 18.90 | 255.56 |
Mean execution time (solid lines) clearly shows that the second optimization significantly outperforms both the original method and the first optimization, nearly halving execution time for small chunk sizes.
Memory usage (dashed lines) stays within a reasonable range across all versions, with only minor increases—highlighting that performance gains were achieved without excessive resource consumption.
Through code refactoring and thoughtful optimization, we reduced execution time by over 50% in some cases with minimal memory overhead, demonstrating measurable and scalable performance improvements.
For this client, our code optimization efforts led to a significant reduction in processing delays, especially in generating complex reports. Users now experience faster results, and the system consumes fewer compute hours - translating to both performance gains and infrastructure cost savings
Dec 2024
Orginal Algorithm
1367 hrs
Jan 2025
Optimized Algorithm
894 hrs
Time saved: ~473 hours
Performance gain: ~35%
Result: Faster reporting
System: Windows Server 2019 Standard Evaluation
DB Engine: MS SQL Server 2022 Express
Processor: QEMU VIRTUAL CPU VERSION 2.5+, 3.79 GHz, 4 CORES / 4 LOGICAL PROCESSORS
RAM: 8GB
DISK: SSD 100GB
NOTE:
The production environment, where the actual time-saving calculations apply, has significantly different hardware and configuration parameters. Test results presented here were collected in a controlled laboratory environment and may vary under live operating conditions
A mid-sized services company was burdened by rising costs from commercial software used for internal communication, project management, and core business operations ⚡
We performed an organization-wide process audit to identify the most cost-intensive areas. As a result, we supported the implementation of open-source alternatives that aligned with the client's needs.
Secure internal messaging
Agile project and issue tracking
Module ERP system for business operations
These tools replaced multiple commercial platforms while preserving functionality and flexibility



Over $2 300/month saved in software and infrastructure licensing



By leveraging trusted open-source technologies, we helped the client build a more cost-effective, adaptable, and self-sufficient software ecosystem
A client operating on a legacy ERP system built with the .NET Framework was facing growing risks: increasing maintenance costs, integration difficulties, and lack of human resources to properly support and develop the system. The aging platform also limited compatibility with modern tools like PrestaShop and new-generation printer services ⚡
We led a strategic migration of the ERP system to .NET 8, enabling long-term support, security, and scalability. Key highlights included

An ERP implementation company working on a multi-year digital transformation project for a client in the air industry needed to quickly scale their programming department ⚡
Websome provided on-demand programming support in the form of a dedicated team (3 FTE) embedded directly into the client's agile process. Supported by client's domain expertise, our developers helped build lots of HR workflows

Our flexible staffing model enabled the ERP partner to meet complex project demands with confidence - without overextending their core team, successfully pushing the project on support and maintenance, as our fruitful cooperation continues!