Marek Kacprzak.
Engineering leader turning ambiguous problems into secure, scalable products — and helping teams ship them.
How I work.
I lead engineering and security for a confidential automation platform, owning the path from an incomplete brief to a reliable production system. That means setting direction, shaping scope, allocating work, reviewing technical decisions, removing blockers, and staying hands-on across APIs, data, cloud infrastructure, security, and operations.
At IBM, I work directly with enterprise customers to uncover the real problem behind a request, turn it into a viable architecture, and prove it through workshops, demos, PoCs, and MVP plans. Earlier delivery and project-management roles taught me to connect roadmap, budget, risk, quality, and stakeholder communication without losing sight of the engineers doing the work.
I am studying Computer Science with a Cybersecurity specialisation at SGGW. I care about teams that ask good questions, make evidence-based trade-offs, document decisions, and improve continuously. My technical range — Python, TypeScript, Go, distributed systems, AI workflows, cloud, and secure delivery — helps me lead with context rather than process for its own sake.
How I create leverage.
- End to end
Product ownership
Roadmap, architecture, delivery, security, releases, and production operations.
- 100 events/s
Proven resilience
6,000-event Kafka verification run completed without failed writes.
- Cross-functional
Team leadership
Backend, frontend, platform, security, operations, product, and customer stakeholders.
- PoC to MVP
Customer translation
Enterprise discovery turned into working proof points and implementation plans.
Systems and products I've led.
Infrastructure Automation Platform
May 2026 — CurrentProduction platform spanning secure builds, artifacts, AI analysis, and resilient delivery.
I lead the roadmap, architecture, delivery standards, security posture, and operational resilience while remaining hands-on across Python/FastAPI services, PostgreSQL workflow state, Kafka distribution, object storage, Kubernetes, GitOps, and private cloud. The platform turns immutable source snapshots into verified artifacts through isolated, observable workflows.
- Python / FastAPI
- PostgreSQL
- Kafka
- Kubernetes / k3s
- OpenTofu
- Ansible
- Argo CD
- Prometheus
Show detailsHide details
A production build, artifact, and infrastructure automation platform. I own the product and engineering lifecycle, coordinating platform, backend, security, and operations while setting the quality bar for releases and reliability.
Leadership & Architecture
- End-to-End Ownership: Set roadmap, architecture, delivery standards, security priorities, and operational practices; translate stakeholder needs into scoped work across engineering disciplines.
- Platform Design: Reproducible private cloud using Proxmox, OpenTofu, Ansible, Kubernetes/k3s, and Argo CD, with Python/FastAPI services for ingestion, builds, artifacts, reporting, notifications, retention, and distribution.
- AI Workflows: AI-driven code and infrastructure analysis with structured outputs, validation, traces, evaluation checks, and human correction paths.
Reliability & Secure Delivery
- Event-Driven Distribution: Kafka with transactional outbox, independent consumers, retries, dead letters, deduplication, and backpressure; verified 6,000 events at 100 events per second without failed writes.
- Release Governance: Workload identity, private PKI, short-lived secrets, SBOMs, provenance, vulnerability scanning, signing, approval gates, verification, rollback, and adversarial testing.
- Operations: Automated CI/CD and GitOps with monitoring, backup, disaster recovery, failure diagnosis, and controlled recovery.
Connexio.pl
Jan 2026 — Apr 2026API-first platform standardising a Transit Connectivity Index for the Warsaw metropolitan area.
Full-stack platform from architecture through deployment — FastAPI over PostgreSQL/PostGIS, Valhalla and OTP routing engines generating isochrones, a Scikit-learn pipeline computing weighted TCI scores, and an Angular frontend with MapLibre GL JS maps and D3.js radar charts. Containerised end-to-end with Redis caching keeping P95 under 300ms on geospatial scoring endpoints.
- Python (FastAPI)
- PostgreSQL / PostGIS
- Angular
- Valhalla / OTP
- Scikit-learn
- MapLibre GL JS
- Redis
- Docker
Show detailsHide details
API-first platform establishing a standardised Transit Connectivity Index (TCI) for the Warsaw metropolitan area. I owned the full lifecycle — from architectural specification to deployment of a containerised data processing engine — prioritising long-term stability and high-performance geospatial analysis.
Technical Architecture & Strategic Planning
- Master Plan Execution: Authored and executed a comprehensive implementation plan, delivering a four-phase roadmap that integrated spatial databases, ML pipelines, and interactive visualisations.
- System Design: Architected a robust stack using Python (FastAPI), PostgreSQL with PostGIS, and Angular, ensuring strict typing and high throughput for both B2C and B2B applications.
- Infrastructure Orchestration: Implemented a fully containerised environment using Docker Compose and Nginx, ensuring strict environment parity and easy portability across cloud providers.
- Performance Optimisation: Configured dedicated Linux infrastructure and Redis caching to achieve P95 response times under 300ms for complex geospatial scoring endpoints.
Development & Operational Excellence
- Data Pipeline Engineering: Developed Python ETL scripts for automated GTFS and OSM data ingestion, maintaining high data integrity through scheduled update cycles.
- Geospatial & ML Logic: Deployed and linked Valhalla / OTP routing engines within Docker to generate accurate isochrones, integrating a Scikit-learn pipeline to compute weighted TCI scores.
- Frontend & Visualisation: Built a responsive Angular application featuring interactive MapLibre GL JS maps and D3.js radar charts for advanced connectivity metrics.
Build Management System
2026 — CurrentSecurity-first build orchestration with secrets handling baked in.
Centralised dashboard, queueing engine, and hardened build server. HMAC-SHA256 request signing, atomic configuration writes, path-traversal validation, automatic secret scrubbing in activity logs, and UFW-isolated traffic for admin-only ranges.
- Node.js
- TypeScript
- Python
- Nginx
- Docker
- HMAC-SHA256
- Linux
Show detailsHide details
Centralised Build Management System and Dashboard orchestrating complex compilation workflows — bridging high-level administrative control with low-level server security to keep a stable environment for automated game builds and real-time infrastructure monitoring.
Full-Stack Dashboard & API Development
- Administrative Interface: Engineered a modern, responsive dashboard with real-time health metrics.
- Management API: Custom management API handling atomic configuration writes, automated backups, and path validation to prevent path-traversal vulnerabilities.
- Build Orchestration: One-click build triggering with advanced concurrency control and a queueing engine for simultaneous resource-intensive tasks.
- Real-Time Monitoring: Services panel reflecting live status of the Build Server API, Discord bots, and Wine environments — immediate visibility into system health.
Infrastructure, Security & DevOps
- Hardened Server Environment: Secured the build infrastructure with UFW rules isolating network traffic to admin-only ranges.
- Security Protocols: HMAC-SHA256 signature system for all requests, with environment sanitisation to protect master secrets from build scripts.
- Logging & Persistence: Resilient logging via .jsonl persistence and automated data scrubbing so sensitive material like PATs and API secrets never surfaces in activity tabs.
- Automated Failure Recovery: Timeout protection and automated build termination logic prevent hung processes from consuming server resources.
Academic Neighbour
Oct 2025 — Nov 2025Real-time social networking platform for dormitory ecosystems.
- Python (FastAPI)
- SQLModel
- PostgreSQL
- Angular
- WebSockets
- Docker
- Traefik
- Oracle Cloud
Show detailsHide details
Real-time social networking platform built for dormitory ecosystems. I focused on a scalable architecture balancing high-performance backend logic with a modern responsive UX, while maintaining a robust and automated infrastructure.
Full-Stack Engineering & Real-Time Systems
- Backend Architecture: Developed a high-performance backend using Python (FastAPI) and SQLModel, implementing complex PostgreSQL schemas and secure JWT-based authentication.
- Frontend Development: Built a modular SPA using Angular v20 with strict typing, Tailwind CSS, and Angular Material for a polished, responsive interface.
- Real-Time Communication: Implemented live chat and dynamic updates by engineering native WebSocket integrations within both FastAPI and Angular services — no Socket.IO shortcut.
- Feature Orchestration: Designed and delivered core platform modules including dynamic content feeds, event management, and administrative profile controls.
Infrastructure & DevOps Excellence
- Cloud Orchestration: Deployed containerised infrastructure on Oracle Cloud (ARM) using Docker Compose and Traefik for automated SSL management and reverse proxying.
- Automated Workflows: Established a modern DevOps pipeline using GitHub Actions for CI/CD, ensuring consistent and reliable feature delivery.
- Development Standardisation: Implemented Devcontainers to give the team a unified, reproducible development environment — eliminating configuration drift.
- Security & Performance: Optimised database queries and container resource allocation to ensure high availability and data integrity across the network.
Headless Browser Scraping Toolkit
Jun 2024 — CurrentGo-orchestrated scraping framework hardened against modern anti-bot defences.
- Go
- Playwright
- Puppeteer
- Redis
- MySQL
- Docker
- Prometheus
Show detailsHide details
Self-directed scraping framework built around Playwright and Puppeteer with a Go control plane — the project where I keep my anti-bot and stealth muscle in shape outside client work. Built for sustained success rates over raw throughput.
Stealth & Anti-Bot Engineering
- Fingerprint Hardening: Browser-profile randomisation covering WebGL, canvas, timezone, locale, and plugin-list signals, tuned against a curated set of known anti-bot fingerprinting probes.
- Stealth Tooling: Integrated and extended stealth plugins for Playwright and Puppeteer, eliminating automation tells that off-the-shelf libraries miss.
- Networking Layer: Residential and datacentre proxy rotation with per-proxy health scoring, per-domain request pacing, and TLS-fingerprint awareness to blend with organic traffic.
- Hostile Targets: CAPTCHA escape hatch and adaptive retry policies for sites that escalate defences mid-session.
Reliability & Operational Excellence
- Worker Orchestration: Go control plane scheduling Playwright / Puppeteer workers behind a Redis-backed queue, with retry / backoff (jitter), per-page timeouts, and screenshot-on-failure for forensic debugging.
- Observability: Prometheus metrics scoped per target — success rate, latency, and proxy health are visible per domain rather than aggregated into one useless number.
- Persistence & Scaling: MySQL for run metadata, Redis for in-flight state, stateless workers so capacity scales horizontally with proxy budget.
- Infrastructure: Fully containerised via Docker for environment parity between local debugging and production runs.
Nest.js Analytics API with BigQuery
May 2025 — Dec 2025Typed, cost-aware analytics endpoints over a warehouse-scale dataset.
- Nest.js
- TypeScript
- BigQuery
- PostgreSQL
- Redis
- Jest
- Playwright
- OpenAPI
Show detailsHide details
Self-directed Nest.js + TypeScript backend exposing typed, cost-aware analytics endpoints over a warehouse-scale BigQuery dataset. I owned the full lifecycle from schema and API design through testing and deployment, prioritising strict typing, query economics, and a defensible test pyramid over feature velocity.
API Design & Architecture
- Modular Nest.js Layout: Controllers, services, and providers with clear boundaries and dependency injection — easy to extend without leaking concerns across layers.
- Strict Typing: TypeScript end-to-end with DTO validation (class-validator) at every request boundary, so malformed inputs are rejected before they touch a query.
- API Contract: OpenAPI specs generated directly from the codebase, keeping the contract in lock-step with the implementation.
BigQuery Query Layer & Performance
- Cost-Aware Queries: Partition-pruning by date, column projection, and bounded pagination so each endpoint has a predictable scan cost rather than an open-ended bill.
- Injection-Resistant: Parameterised queries with type-checked bindings — no string interpolation into SQL.
- Redis Caching: Result caching with per-endpoint TTLs and cache keys derived from validated DTOs, so hot dashboards stop re-billing BigQuery for identical reads.
Testing & Operability
- Test Pyramid: Jest unit tests for service logic, supertest integration tests against a Dockerised PostgreSQL for metadata, and a Playwright suite covering auth and dashboard end-to-end.
- Containerised: Docker Compose for local parity with deployed environments; structured logging for traceable production debugging.
Passkey Wallet
Aug 2025 — Feb 2026ERC-4337 v0.7 smart wallet with WebAuthn passkey signatures and a sponsor paymaster.
- Solidity
- ERC-4337
- Foundry
- WebAuthn
- Next.js
- viem
- Hono
- Playwright
Show detailsHide details
ERC-4337 v0.7 Passkey Wallet — a smart wallet that accepts an EOA owner (secp256k1) or a WebAuthn passkey (secp256r1), with gas paid by an off-chain-signed sponsor paymaster. The Coinbase Smart Wallet pattern, built end-to-end.
ERC-4337 v0.7 Contracts
- PasskeyAccount: BaseAccount-derived, dispatches signature validation to either an ECDSA path or a WebAuthn path that calls the RIP-7212 P-256 precompile with a fallback verifier. UUPS-upgradable, gated to address(this) or the EntryPoint.
- PasskeyAccountFactory: CREATE2 + ERC-1967 proxy. Idempotent — a second call with the same salt returns the same proxy without reverting.
- SponsorPaymaster: Verifying paymaster signed under (validUntil, validAfter); digest bound to chainid + paymaster address to block cross-chain / cross-paymaster replay.
- Test Coverage: 11 Foundry tests via real entryPoint.handleOps calls, asserting AA24 / AA32 / AA34 EntryPoint reverts for the failure paths.
Off-Chain Sponsor Service
- Hono Server: Holds the sponsor key; allowlist policy is fail-closed (empty allowlist → 403). Authoritatively decides what gets sponsored, with no implicit defaults.
- Signature Layout: Signs paymasterAndData in the v0.7 layout (paymaster | gas limits | validity | sig) — sign.ts mirrors SponsorPaymaster.getHash() exactly so any drift surfaces in on-chain verification, not in production.
- API Documentation: Hand-written OpenAPI 3.0 spec, Redocly-linted in CI; 4 vitest tests covering signing, allowlist, fail-closed, and malformed-body paths.
Next.js Dashboard
- Wallet Stack: wagmi v2 + RainbowKit + viem v2. Counterfactual address derivation via factory.getAddress (view call, free).
- WebAuthn Enrolment: navigator.credentials.create() with ES256, exporting the P-256 (x, y) the on-chain account stores.
- E2E Testing: 10 Playwright tests including passkey enrolment driven through Chromium's virtual WebAuthn authenticator via CDP — proves the full enrolment → public-key-export round-trip, not just stubs.
Multi-Chain DApp
Jan 2026 — CurrentWeb3 onboarding and NFT marketplace across EVM, Sui, and Solana.
- Next.js
- TypeScript
- Solidity
- Move
- ethers.js
- viem
- Foundry
- The Graph
- IPFS
Show detailsHide details
Self-directed multi-chain DApp applying production-grade web dev and security discipline, with blockchain as the new integration surface.
Frontend & Wallet Connector Layer
- Type-Safe Stack: Next.js + TypeScript frontend with responsive UX patterns lifted from production Angular / React work.
- Modular Wallet Connectors: Unified connector layer across MetaMask, Phantom, WalletConnect, and Sui Wallet — chain-aware routing without hardcoding wallet-specific quirks at the call site.
Smart Contracts & On-Chain Logic
- EVM: Solidity ERC-721 contracts via ethers.js and viem; The Graph for event indexing.
- Sui: Move modules using @mysten/sui.js for object-centric NFT logic.
- Solana: Explored Solana program calls via @solana/web3.js, covering wallet connection, devnet transfers, and account-state reads.
- Off-Chain Storage: IPFS (Pinata) for NFT metadata, with Arweave evaluated as the permanence alternative.
Security as First-Class
- Signature Verification: SIWE-style signature verification on authenticated actions with replay-protected nonces.
- Hardened RPC Proxy: Rate-limited RPC proxy with HMAC-signed internal APIs — DevSecOps patterns transferred directly from professional infrastructure work.
Testing & Scalability
- Contract Tests: Foundry contract unit tests plus signing-flow integration tests covering authenticated paths.
- Backend Contract: OpenAPI-documented backend routes; layered architecture that scales horizontally behind the RPC proxy.
Smart Contract Security Lab
Jun 2025 — Mar 2026Reproducible exploit PoCs across canonical Solidity vulnerability classes.
- Solidity
- Move
- Slither
- Foundry
- Hardhat
- TypeScript
- Next.js
Show detailsHide details
A catalog of intentionally-vulnerable Solidity contracts paired with verified fixes, static analysis, and a wallet-authenticated PoC dashboard. Every bug is reproducible and every patch provably effective.
Vulnerability Catalog & Exploit Engineering
- Five Classes: Paired Vulnerable.sol / Fixed.sol for reentrancy, signature replay, access control (missing modifier + tx.origin), unchecked-math underflow, and single-block oracle manipulation — modelled on real incidents.
- Differential Foundry Suite: 16 tests where exploits prove the bug and fixes prove the patch closes it — CEI, EIP-712 digest binding, Ownable2Step, checked math, anchor-bounded oracle.
- Static Analysis: Automated Slither runs with curated remappings; committed JSON findings ship with the repo.
- Sui Move Contrast: A safe_vault module showing how Move's linear-resource type system makes several Solidity bug classes structurally impossible.
Dashboard & SIWE Attestation
- Catalog UI: Next.js 14 App Router dashboard hydrating each detail page from contracts/ — live source of both implementations alongside Slither output.
- SIWE PoC API: Two-step EIP-4361 flow — /challenge issues single-use nonces bound to slug + address; /submit verifies via siwe, confirms the signer, records the submission.
- Wallet Integration: wagmi v2 + viem v2 + RainbowKit for MetaMask / WalletConnect.
- API Contract: Hand-written OpenAPI 3.0 spec, Redocly-linted in CI.
End-to-End Test Automation
- Playwright Suite: 13 tests covering catalog UI, API, and a full SIWE round-trip with a real viem-signed message — including a nonce-replay rejection test.
- Verification Gate: forge test (16/16), sui move test (2/2), Redocly lint, npm run build, and Playwright all wired into the documented quickstart.
Glioma Classification
Mar 2025 — Jun 2025Biomedical ML pipeline replicating a 2022 NIH glioma grading methodology from clinical and molecular data.
- Python
- scikit-learn
- PyTorch
- Ensemble Learning
- Feature Selection
- pandas
- Biomedical ML
Show detailsHide details
R&D project replicating a biomedical study for glioma grading using clinical and molecular data. Acted as the bridge between research objectives and technical execution — keeping the ML workflow aligned with rigorous scientific standards and NIH-based methodologies.
Research & Development
- Study Replication: Directed the end-to-end project lifecycle for glioma grading, replicating the source biomedical study through precise feature selection and model evaluation.
- Methodology Compliance: Reviewed project documentation against the 2022 NIH-based reference methodology and scientific protocols to keep results defensible.
- Technical Coordination: Scheduled and oversaw specialised tasks involving ensemble learning (Soft Voting) and predictive modelling for high-fidelity results.
- Specialist Collaboration: Facilitated continuous communication and technical alignment between machine learning specialists and data researchers throughout the study.
Project Governance & Stakeholder Alignment
- Lifecycle Planning: Managed the project roadmap in Jira and Miro, translating research milestones into actionable development tasks and technical sprints.
- Quality Assurance & Documentation: Maintained comprehensive knowledge bases so all research findings and technical implementations were documented for reproducibility.
- Resource & Risk Monitoring: Identified potential blockers in the data processing and modelling phases before they derailed the timeline.
Sky Clustering
Mar 2025 — Jun 2025Computer vision pipeline clustering high-volume frame data from vehicle-mounted cameras.
- Python
- OpenCV
- NumPy
- Clustering
- Matplotlib
- Plotly
- Computer Vision
Show detailsHide details
Multidisciplinary computer vision project analysing video telemetry from vehicle-mounted cameras. Bridged data-science experiments and structured project delivery so technical breakthroughs translated cleanly into reports for academic and external stakeholders.
Technical Project Control & Agile Leadership
- Sprint & Milestone Management: Directed the project lifecycle using Agile methodologies, planning sprints and tracking milestones across data preprocessing and cluster analysis phases.
- Workflow Optimisation: Identified and resolved technical bottlenecks so the pipeline for analysing high-volume video frame data stayed seamless.
- Tooling & Infrastructure: Introduced and managed the technical ecosystem — Jira, GitHub, and ProjectLibre — to maintain structured collaboration.
- Resource Coordination: Led a diverse team of specialists, aligning their efforts with strategic goals and ensuring optimal task distribution throughout the research cycle.
Stakeholder Management & Technical Reporting
- Results Visualisation: Translated complex clustering experiment results into clear, data-driven visualisations for presentation to external academic stakeholders.
- Documentation Excellence: Established rigorous documentation standards so all research findings and technical configurations were archived and reproducible.
- Quality Oversight: Monitored the integrity of data analysis workflows, ensuring results met the standards expected by academic partners and industry collaborators.
Academic Research Database Comparator
Oct 2024 — Feb 2025Centralised platform comparing research outputs across Polish university databases.
- Python (FastAPI)
- PostgreSQL
- Angular
- Web Scraping
- Data Integration
- Docker
- GitHub Actions
Show detailsHide details
Centralised platform comparing academic research outputs across various Polish university databases. Synchronised technical workflows with academic requirements through structured team leadership and transparent stakeholder management.
Project Governance & Team Leadership
- Workflow Organisation: Directed a multidisciplinary team by establishing clear operational frameworks, running regular sync meetings, and managing the end-to-end project lifecycle.
- Strategic Task Planning: Used Jira to plan and track complex tasks, keeping development aligned with the project roadmap and technical specifications.
- Specialised Delegation: Assessed team members' technical strengths and delegated tasks accordingly, providing mentorship to maintain high standards of execution.
- Resource Management: Used ProjectLibre and Miro to create detailed timelines and visualise project architecture, preventing bottlenecks in the data integration phases.
Stakeholder Management & Delivery
- Academic Alignment: Acted as the primary liaison for university stakeholders, making sure the platform met the rigorous data comparison standards required by academic institutions.
- Documentation & Reporting: Authored and maintained comprehensive project documentation as a clear audit trail of technical decisions and platform capabilities.
- Presentation of Results: Delivered final project outcomes and platform demonstrations to academic boards, highlighting the tool's impact on research visibility and data accessibility.
A career built across engineering and delivery.
Confidential Client
Remote · WarsawHead of Software Engineering
Jan 2026 — CurrentOwn the product and engineering lifecycle for an AI-enabled automation platform. Set roadmap and technical direction, allocate work across backend, frontend, platform, security, and operations, review architecture and delivery quality, remove blockers, and remain hands-on in Python/FastAPI, event-driven workflows, cloud infrastructure, observability, and secure releases.
Head of Software Development and Security
Jan 2026 — Jun 2026Established the platform architecture and security strategy across multi-cloud infrastructure, CI/CD, access control, monitoring, and AI-driven analysis. Built the engineering foundations for reproducible delivery, operational resilience, and protection of source code and proprietary systems.
IBM
Full-time · Warsaw — HybridBrand Technical Specialist
Jan 2026 — CurrentLead technical discovery for enterprise opportunities and turn ambiguous customer needs into viable solution architectures, demos, proofs of technology, PoCs, workshops, proposals, and MVP plans. Coordinate customers, engineering, product, architecture, and sales while communicating constraints, risk, scalability, and business value.
Pre-sales Engineer
Oct 2025 — Dec 2025Acted as the technical bridge between product capabilities and enterprise requirements, owning solution design, customer workshops, technical proof points, integration guidance, and RFP/RFI workstreams.
Software Mind
Full-time · Warsaw — HybridJunior Delivery Manager
Jun 2025 — Sep 2025Managed cross-functional delivery across fintech, SaaS, and public-sector work: requirements, Agile execution, risks, SLAs, quality, testing, deployment support, budgets, and PMO reporting. Coached team members, improved onboarding and feedback loops, and automated internal operational workflows.
Weblo
Full-time · Warsaw — On-siteProject Manager
Apr 2025 — Jun 2025Led e-commerce projects from scoping and requirements through planning, risk management, design-to-code quality assurance, documentation, and deployment. Served as the primary client contact and translated commercial priorities into actionable technical work.
OSEC — Open Source Education Center
Part-time · Warsaw — HybridIT Project Manager
Apr 2024 — Jun 2025Led delivery of data-driven products for the Allegro seller ecosystem across roadmap, backlog, sprints, risks, documentation, quality reviews, testing, and cloud releases. Coordinated stakeholders, UX, developers, and testers using Jira and Azure DevOps for transparent project reporting.
Full Stack Developer
Jan 2024 — Apr 2025Built production analytics software across Go and Node.js APIs, PostgreSQL, Angular, Tailwind, and a WXT browser extension. Operated Docker and Kubernetes environments, CI/CD, automated tests, monitoring, and production debugging — technical depth that still shapes how I estimate and lead delivery.
Concepcion Ideale N2
Full-time · Warsaw — RemoteFull Stack Developer
Jun 2022 — Oct 2023Built and deployed a secure document-purchasing and electronic-signature platform using Node.js, Express, MySQL, JavaScript, DigitalOcean, and CI/CD. Owned database design, backend workflows, responsive UI, secure transaction handling, releases, and performance improvements.
How I lead and build.
- Leadership & Management
- Technical direction · Roadmaps & prioritisation · Work allocation · Mentoring & feedback · Risk & budget management · Quality governance
- Product & Stakeholders
- Customer discovery · Problem framing · PoCs & MVPs · Cross-functional facilitation · RFP / RFI · Executive communication
- Systems & Backend
- Python / FastAPI · Node.js / TypeScript · Go · REST / OpenAPI · Event-driven architecture · Distributed systems
- Data & AI
- PostgreSQL / PostGIS · BigQuery · Redis · Kafka · ETL & ingestion · LLM & agent workflows
- Cloud & Reliability
- AWS / GCP / Azure · Kubernetes · Docker · Terraform / OpenTofu · Ansible / Argo CD · Prometheus / Grafana
- Delivery & Security
- Agile / Scrum · CI/CD & GitOps · Automated testing · Threat modelling · Secure SDLC · Incident recovery
Build something that matters.
I'm interested in engineering leadership, technical management, and senior product-engineering roles where I can help a strong team turn complex problems into dependable systems. The best way to reach me is email.
- [email protected]
- GitHub
- navitelogger
- marek-kacprzak
- Based in
- Warsaw, Poland