15+ years of commercial experience building and maintaining production web systems. For the past years I've specialized in applied AI: LLM agents, real-time voice AI, RAG, and NLP with personal-data anonymization.
Sole author of AI services running in production: job-title classification to occupation codes (Azure OpenAI, vector search, RAG) and a complete CV parsing & PII-anonymization pipeline (Presidio, NER - spaCy/Stanza, OCR).
Evolved a recruitment voicebot from a scripted IVR to natural speech-to-speech conversations on the OpenAI Realtime API and ElevenLabs Conversational AI (Twilio Media Streams, barge-in handling).
~2,600 PHP files, custom MVC framework, ~100 job-board integrations, SOLR search. Recovered and optimized its MySQL performance three times under production load.
Multilingual voice agent (11 languages) on Gemini 2.5/3.5 Flash: versioned prompts, anti-prompt-injection and GDPR guardrails, automated deployment pipeline.
Candidate-to-offer matching engine (Assistants API + Vector Stores) with statistical quality validation against recruiter judgments (Spearman correlation).
PhD studies in automation & robotics (time series, RNN/LSTM, sensor fusion) and 15+ publications, incl. Scientific Reports and Oncotarget.
🌱 Currently exploring: Brave Search API - web-search tooling for LLM agents.
PhD studies in automation & robotics (Opole University of Technology, 2015–2019; programme discontinued): time-series modeling with recurrent neural networks (RNN/LSTM), sensor fusion and adaptive filtering for mobile-device motion tracking (MEMS sensors). Author/co-author of 15+ scientific and conference publications, incl.:
No. DS-SD/72/WE/17, Opole University of Technology - "Positioning of mobile devices using adaptive-filtering tools and identification methods based on MEMS sensors."
Looking for an experienced Fullstack / AI developer? I'd be happy to talk.