softmedia/ digital systems
PL
System online00:00:00
ROOT / ABOUTSECURE CONNECTION
PROFILE / PIOTR_SZYMAŃSKI
Piotr Szymański - CIO, systems architect and founder of SoftMedia
INFO / ABOUTCIO · ENTERPRISE ARCHITECT · SOLUTION BUILDER

From IT strategy to a working system.

SoftMedia is led by Piotr Szymański, a CIO, systems architect and software developer. We combine more than 20 years of technology experience with business processes, from sales systems and ERP through integrations and data platforms to AI, RAG and agent architecture.

SoftMedia has operated since 2007. Piotr participates personally in process analysis, application and data architecture, code review, Proof of Concept work and production delivery. An executive role has never taken him away from technology.

01 / PROFILE

From software developer to CIO, while staying close to technology.

Piotr Szymański began his career in 2005 as an enterprise software developer. One of his first major projects was a call-centre system designed from the ground up, which still supports more than 70 consultants.

In subsequent years he took responsibility for infrastructure, cybersecurity, software and administration teams, sales systems and digital transformation. Since 2020, as Chief Information Officer, he has led technology development at Forum Media Polska, part of an international group operating on four continents.

Between 2022 and 2024 he was also responsible for building the controlling function. This experience adds processes, cost, KPIs, risk, profitability and management information quality to his technology perspective.

This path connects the viewpoints of users, business, architecture and technical delivery. Piotr understands not only how to build a system, but why an organisation needs it, how to deploy it responsibly and how to preserve its value.

CAREER_TIMELINEEngagement timeline
  1. Software DeveloperWydawnictwo FORUM
    20052013
  2. Owner · Solution ArchitectSoftMedia
    2007presentACTIVE
  3. IT ManagerForum Media Polska
    20132020
  4. Chief Information OfficerForum Media Polska
    03.2020presentACTIVE
  5. Head of ControllingForum Media Polska · parallel role
    20222024
02 / OPERATING_MODEL

We connect perspectives that are usually separated.

We move from business goals, risk and budgets to data models, API contracts and concrete implementation decisions. Less information is lost between business, architecture and delivery.

BUSINESS / STRATEGY

Direction, accountability and business value

IT strategy, project portfolios, governance, cybersecurity, board-level cooperation, change management and assessment of technology investments.

ARCHITECTURE / DATA

Process, systems and information

Process analysis, application architecture, data models, ERP, CRM, MES, integrations, permissions, security and enterprise solution design.

DELIVERY / CODE

From a technical decision to production

Proofs of Concept, MVPs, implementation decisions, code review, testing, DEV and PROD environments, monitoring, deployment and iterative product development.

03 / FOCUS_AREAS

Difficult problems where processes, data and software meet.

We do not begin with a fashionable technology. We first understand the process, the data, the decisions and the exceptional cases. Architecture and tools follow.

AI_EMBEDDED_IN_BUSINESS

AI embedded in business processes

We design solutions using LLMs, RAG, agent architecture, MCP and document analysis. Models work with controlled context, respect permissions, cite sources and leave an auditable trail. Sensitive data can be handled through local, hybrid or on-premises architectures.

ENTERPRISE_SYSTEMS

Systems designed around real work

We build custom business applications, ERP, CRM and MES solutions supporting sales, customer service, manufacturing, finance and logistics, with close attention to domain models, integration, scalability and maintainability.

DATA_AND_DECISIONS

Data that people can trust

Our work covers data quality, normalisation, deduplication, migration, database architecture, reporting, forecasting and machine learning. Systems should help assess information quality while keeping people in control of ambiguous decisions.

AUTOMATION_AND_INTEGRATION

Automation designed for exceptions

We connect systems using REST APIs, webhooks, n8n, asynchronous communication and custom components. Reliable automation includes validation, least privilege, idempotency, monitoring, retries and human approval for higher-risk actions.

04 / SELECTED_PROJECTS

Projects that best demonstrate how we work.

Each combines process analysis, architecture, data and responsibility for bringing a solution into real use.

LEXCOREAILEGALTECH + DOCUMENT AI

AI platform for legal operations and large case-file analysis

A complete business application combining document management, court-system integration and verifiable AI analysis. Models work on a defined material version, respect permissions and can cite the documents and pages used in an answer.

RESPONSIBILITY

Solution concept, application and data architecture, AI workflows, integration, technology selection, security and product development.

.NETPYTHONRAGAI AGENTSDOCUMENT AIDOCKER
View case study: LexCoreAI
DYNAMICS_365_BCERP TRANSFORMATION

ERP transformation in an international organisation

Strategic professional experience covering process analysis and optimisation, solution architecture, Microsoft Dynamics 365 Business Central extensions, API integration, backlog management and production readiness.

RESPONSIBILITY

One of the programme’s business and technology leaders, connecting users, the implementation partner, the IT team and group headquarters.

BUSINESS CENTRALERPENTERPRISE ARCHITECTUREAPICHANGE
STRATEGIC EXPERIENCE / PROFESSIONAL ROLE
DBFIXSTUDIODATA QUALITY + ML

Deduplication and normalisation of long-running customer databases

A proprietary platform turns one-off database clean-up into a controlled data-quality process. Normalisation rules, similarity methods and ML.NET reduce manual work while retaining source data, operator decisions and a complete audit trail.

RESPONSIBILITY

Solution architecture, normalisation engine, deduplication workflow, machine-learning layer and review interface.

.NETSQLML.NETDATA QUALITYHUMAN IN THE LOOP
View case study: DbFixStudio
OPENMESPRODUCTION + PWA

An MES designed around real operator work

Co-development and pilot deployment of an open-source project supporting planning, production batches, materials, quality, waste and downtime. The workstation interface is designed for tablet use under real shop-floor conditions.

RESPONSIBILITY

Process analysis, new feature design, operator experience, pilot configuration, custom extensions and integrations.

MESPWAPRODUCTIONUXINTEGRATIONS
View case study: OpenMES
View all case studies
05 / TECHNOLOGY_STACK

Technology is a tool, not the objective.

Our deepest experience is in C#, .NET and SQL Server. AI and analytics projects also use Python, FastAPI, PostgreSQL, machine-learning models and vector search, deployed in containerised environments.

We validate new technologies through practical use and select them when they fit project requirements, not merely because they are popular.

CORE_SOFTWARE

  • C#
  • .NET
  • ASP.NET Core
  • SQL Server
  • REST API
  • Software Architecture
  • Database Design

AI_AND_DATA

  • Python
  • FastAPI
  • PostgreSQL
  • TimescaleDB
  • LLM
  • RAG
  • Vector Search
  • AI Agents
  • MCP
  • ML.NET
  • XGBoost
  • LightGBM

PLATFORM_AND_AUTOMATION

  • Docker
  • Portainer
  • Linux
  • Redis
  • Nginx
  • Git
  • PowerShell
  • n8n
  • Webhooks
  • Monitoring

BUSINESS_SYSTEMS

  • Dynamics 365 Business Central
  • ERP
  • CRM
  • MES
  • PWA
  • System Integration
  • Business Process Automation
06 / WORK_PROTOCOL

We start with the process, not the technology.

  1. 01 / DISCOVERY

    Understand the problem

    We speak with process owners and users to establish goals, data sources, constraints, exceptions and the criteria used to evaluate the result.

  2. 02 / ARCHITECTURE

    Design the solution

    We design the target process, data model, system boundaries, integrations, permissions, security, error handling and maintainability.

  3. 03 / PILOT

    Validate assumptions with a focused scope

    A Proof of Concept, MVP or pilot validates the most important risks early and puts the solution in front of real users.

  4. 04 / DELIVERY

    Deploy to production

    We prepare tests, environments, monitoring, data migration and go-live. The system must work in everyday operations, not only during a presentation.

  5. 05 / DEVELOPMENT

    Improve through real use

    After launch we analyse outcomes, defects and user feedback. Further features are driven by observed value rather than initial assumptions alone.

07 / DEVELOPMENT_MINDSET

Experience does not mean attachment to old solutions.

We treat more than 20 years of experience as a foundation, not a reason to remain with familiar technology. We continue to prototype, evaluate new models, compare approaches, work with code and test how technology behaves beyond a demonstration.

Our current focus includes LLMs, RAG, AI agents, MCP, document analysis and architectures for confidential information. New technology must solve a real problem, fit the client’s infrastructure and remain secure and maintainable.

A management perspective helps us see the wider organisation. Active technical work keeps us close to what actually works.

08 / FAQ

Common questions about working with us.

Q_01

Does Piotr Szymański participate personally in projects?

Yes. Piotr is personally involved in process analysis, architecture, technology selection, implementation decisions, solution review and production preparation. His CIO role does not mean managing delivery from a distance.

Q_02

Does SoftMedia work exclusively on AI projects?

No. AI is one of our active areas. We also deliver custom business applications, ERP, CRM and MES systems, integrations, data platforms, automation and reporting solutions.

Q_03

Can an engagement begin with a small scope?

Yes. We can start with one process, an architecture review, a Proof of Concept, an MVP or a pilot with a limited user group.

Q_04

Can an AI solution run locally or on premises?

Yes, when required by data confidentiality, regulation or infrastructure. The architecture and model are selected according to quality, cost, performance, integration and security requirements.

Q_05

Do you work only with Microsoft technologies?

Our deepest experience is in C#, .NET and SQL Server, but technology follows the project. Current solutions also use Python, PostgreSQL, FastAPI, Docker, Linux, n8n and AI or machine-learning technologies.

READY_FOR_DISCOVERY

Have a process that off-the-shelf software cannot support?

We can begin with one problem, an architecture review, a pilot or a conversation about combining software, data and AI.