KAMSOFT

Data Science · Data Engineering · Scientific Computing

DATA · MODELS · INSIGHTS

Independent consulting for research-driven companies and individual clients. I turn exploratory code, numerical models, and ad-hoc data workflows into reliable, maintainable systems.

Quick facts

• Owner: Kamil Bruchal (Kamsoft)

• Email: kamsoft@kambr.pl

• Website: kambr.pl/kamsoft

• Focus: Data Science · Data Engineering · Scientific Computing

• Working style: correctness, reproducibility, maintainability

About

I come from a scientific and computational background, with experience working at the boundary between research and engineering. Much of my work involves taking exploratory code, numerical models, or ad-hoc data workflows and turning them into reliable, maintainable software systems.

In addition to R&D-focused projects, I am available for general data-oriented and software development work, including internal tools, backend systems, and improvements to existing codebases—especially where performance, correctness, or technical debt are concerns.

I work directly with clients, keep assumptions explicit, and aim to deliver solutions that can be understood and maintained without ongoing external dependency.

Typical Problems I Solve

• Research or prototype code that cannot be scaled or maintained

• Data pipelines that rely on manual steps or undocumented scripts

• Numerical models that are slow, unstable, or difficult to validate

• Machine learning projects without reproducibility or a clear deployment path

• Inefficient or fragile HPC or cloud compute setups

• R&D results that cannot be cleanly transferred into production systems

Expertise Areas

Data Engineering

• Design and implementation of reliable data pipelines

• ETL/ELT processes and data warehousing

• Cloud and hybrid data architectures

• Batch and streaming data processing systems

Data Science

• Statistical analysis and predictive modeling

• Machine learning workflows with reproducibility and evaluation discipline

• Feature engineering, dataset construction, and experiment tracking

• Model deployment integration (when needed)

Scientific Computing

• Numerical simulations and model implementation

• Performance optimization (vectorization, parallelism, profiling)

• High-performance computing workflows

• Reproducible computational research tooling

Software Development

• Backend systems and internal tooling

• API design and automation

• Codebase refactoring and technical debt reduction

• Testing and CI practices for scientific / data-heavy code

Tools & Stack (typical)

• Python, Julia

• NumPy/SciPy, Pandas, Polars

• SQL, DuckDB, Postgres

• Docker, CI/CD

• HPC environments, Linux tooling

Engagement

I typically work in one of these modes:

• Audit / assessment: identify failure modes, bottlenecks, and a realistic roadmap

• Implementation: build or refactor the system, with tests and documentation

• Support: short-term help for a specific hard problem (performance, correctness, migration)

Get in touch

If your data, models, or compute setup is becoming a bottleneck, email me:

kamsoft@kambr.pl

Or visit: kambr.pl/kamsoft

Kamsoft — Kamil Bruchal NIP: 8993048156 · REGON: 543855110 · Email: kamsoft@kambr.pl