Staff Product Engineer · Agentic AI Workflows · Privacy-First AI Infrastructure · Design Systems

I ship production systems end-to-end — from front-end architecture and design systems to local-first AI infrastructure. Currently on contract with a major tech company, building a new platform where agentic coding is the daily workflow — orchestrating coding agents at serious token budgets, owning features end-to-end across web and native surfaces. 15+ years of professional experience (CGI, KPMG, Acxiom). Solo-built a subscription SaaS and a privacy-first AI tool ecosystem.

Open to founding / staff engineering roles, technical co-founder opportunities, and selective consulting — Q4 2026 availability.

→ awolczuk [at] gmail [dot] com · LinkedIn · Warsaw · Remote

Building a greenfield platform for a major tech company with agentic coding as the core delivery model:

• Orchestrating coding agents end-to-end — architecture, agent direction, review, and quality gates on agent output
• Token budgets of hundreds of dollars per engineer per day — production-scale agentic development, not experimentation
• Full features shipped across web and native surfaces, from design-system components to complete integrations

Architected a modern insurance PWA achieving a 96% Google Page Speed score:

• Component library with full Storybook documentation and testing
• Storyblok integration enabling visual editing independent of dev cycles
• AI-augmented delivery pipeline using Figma Code Connect to sync design tokens with production code

Full-stack subscription web app shipped to production. Stack highlights:

• Streaming LLM responses — Vercel AI SDK + Claude, token-by-token rendering
• Custom auth bridge — Clerk JWT decoded inside Supabase RLS policies (non-standard integration)
• Heavy compute in serverless — C library compiled to WASM, singleton-cached across requests
• Subscription billing — Stripe Checkout with dynamic pricing and promo codes

One philosophy across all repos: AI should work for you, not harvest your data. Local inference.

pseudonym-mcp — replaces PII with reversible tokens ([PERSON:1], [PESEL:1], [CREDIT_CARD:1]) before your prompt reaches the cloud, then restores original values in the response. GDPR-compliant pseudonymisation, fully offline, works with Claude, GPT-4, Gemini and any MCP-compatible client. Early-stage with an active roadmap — the direction is a deterministic sanitization layer for any agent workflow touching sensitive data.

claude mcp add pseudonym-mcp -- npx pseudonym-mcp

Node.js Apple Vision ecosystem for local OCR & image analysis on macOS — same engine, different surfaces:

• macos-vision — the foundation. Apple Vision OCR for Node.js, native and fast.
• macos-vision-mcp — MCP server for Claude Code & Claude Desktop. Pre-extracts text before your AI sees it, cutting token usage by ~97% on real documents. Also does on-device UI testing: screenshots, accessibility trees, and element lookup — nothing leaves the machine. Built on the newest Apple Vision capabilities: live web-page operation for agents, fully on-device. Combined with pseudonym-mcp, this raises the ceiling on what an agent can safely automate on the web.
• sortai — macOS CLI that OCRs every PDF and image in a folder with Apple Vision and writes native Finder tags (#Faktura, #Umowa, #Bank…) and comments into file metadata — instantly searchable in Spotlight, no database. Offline via Ollama by default; cloud LLMs optional, with pseudonym-mcp masking PII first.

claude mcp add macos-vision-mcp -- npx macos-vision-mcp

sententim — deterministic MCP server that verifies whether a Polish court judgment signature actually exists before you cite it. LLMs hallucinate legal case numbers; this tool makes them stop. Zero LLM at runtime — purely SQLite + FTS5 lookups, sub-10ms, fully offline. Returns FOUND, NOT_FOUND, or AMBIGUOUS against a local database built from SAOS public data.

• Your Obsidian Vault as a Private Second Brain — Powered by Local AI — how to use macos-vision-mcp + pseudonym-mcp to query your vault with cloud AI without exposing personal data.
• Your Messaging Apps as a Private Document AI — Powered by OpenClaw — using macos-vision-mcp + pseudonym-mcp inside OpenClaw to OCR and anonymise documents sent over WhatsApp, Telegram, or Slack before they reach any cloud LLM.
• Apple Vision vs Tesseract — A 50-File OCR-to-Markdown Benchmark — head-to-head OCR comparison on 50 PDFs with identical LLM formatter input; Tesseract wins on CER, Apple Vision wins on structural quality.
• Privacy Tiers for Document AI — Three Pipeline Configurations — fully local vs local OCR + cloud LLM vs local OCR + pseudonymisation + cloud LLM: what each configuration protects, where the trade-offs are, and how to choose.
• LLM as a Bridge Between Qualitative and Quantitative Research — using LLM-as-coder and LLM-as-judge to classify interviews and project free-form text onto validated scales (PHQ-9, BDI-II, SF-36), with bias, non-determinism, and IRB caveats.

Outside software I serve on the management board of a large residential community in Warsaw, co-managing a 7-figure budget and vendor contracts.