{"id":9,"date":"2026-01-31T17:35:52","date_gmt":"2026-01-31T17:35:52","guid":{"rendered":"https:\/\/michaelprinc.com\/?page_id=9"},"modified":"2026-03-30T21:45:33","modified_gmt":"2026-03-30T21:45:33","slug":"uvod","status":"publish","type":"page","link":"https:\/\/michaelprinc.com\/","title":{"rendered":"Michael Princ – Data Scientist & ML Engineer"},"content":{"rendered":"

Pom\u00e1h\u00e1m firm\u00e1m a jednotlivc\u016fm vyu\u017e\u00edt s\u00edlu dat a um\u011bl\u00e9 inteligence pro lep\u0161\u00ed rozhodov\u00e1n\u00ed a efektivn\u011bj\u0161\u00ed pr\u00e1ci.<\/p>\n

\n
\n Zjist\u011bte, jak v\u00e1m mohu pomoci<\/a>\n <\/div>\n
\n Kontaktujte m\u011b<\/a>\n <\/div>\n<\/div>\n
\n

Kdo jsem a co d\u011bl\u00e1m pro v\u00e1s<\/h2>\n

Jsem Michael Princ<\/strong> \u2013 Data Scientist a ML Engineer z Prahy. Kombinuji 9 let akademick\u00fdch zku\u0161enost\u00ed<\/strong> s prax\u00ed z bankovnictv\u00ed, ve\u0159ejn\u00e9ho sektoru a freelance konzultac\u00ed. Rozum\u00edm nejen technologi\u00edm, ale i byznysov\u00e9mu kontextu \u2013 a dok\u00e1\u017eu slo\u017eit\u00e9 koncepty vysv\u011btlit srozumiteln\u011b.<\/p>\n

Pom\u00e1h\u00e1m firm\u00e1m i jednotlivc\u016fm pochopit, jak data a AI mohou konkr\u00e9tn\u011b zlep\u0161it jejich pr\u00e1ci. Neprod\u00e1v\u00e1m obecn\u00e9 rady \u2013 dod\u00e1v\u00e1m funk\u010dn\u00ed \u0159e\u0161en\u00ed a sd\u00edl\u00edm zku\u0161enosti z des\u00edtek re\u00e1ln\u00fdch projekt\u016f.<\/p>\n

\n

Data Science, ML modely a konzultace<\/h2>\n

Nab\u00edz\u00edm komplexn\u00ed datov\u00e9 slu\u017eby \u2013 od anal\u00fdzy dat p\u0159es v\u00fdvoj ML model\u016f a\u017e po jejich nasazen\u00ed do produkce.<\/p>\n

\ud83d\udd2c Data Science konzultace<\/h3>\n

Anal\u00fdza dat, prediktivn\u00ed modelov\u00e1n\u00ed a vizualizace pro pochopen\u00ed va\u0161eho byznysu. Pomohu v\u00e1m odhalit vzory v datech a transformovat je na konkr\u00e9tn\u00ed obchodn\u00ed rozhodnut\u00ed.<\/p>\n

\ud83e\udd16 Machine Learning \u0159e\u0161en\u00ed a ML modely<\/h3>\n

V\u00fdvoj ML model\u016f a end-to-end \u0159e\u0161en\u00ed pro automatizaci a optimalizaci proces\u016f. Od n\u00e1vrhu p\u0159es tr\u00e9nink a\u017e po nasazen\u00ed do produkce. Specialuji se na prediktivn\u00ed modely, scoring, MLOps a Databricks pipelines.<\/p>\n

\ud83c\udf93 Praktick\u00e9 workshopy o AI<\/h3>\n

Praktick\u00e9 workshopy a \u0161kolen\u00ed v oblasti AI a datov\u00e9 analytiky. Nau\u010d\u00edte se vyu\u017e\u00edvat modern\u00ed technologie pro va\u0161i pr\u00e1ci a projekty. Workshopy prob\u00edhaj\u00ed v mal\u00fdch skupin\u00e1ch \u2013 d\u016fraz je na okam\u017eitou pou\u017eitelnost, ne such\u00e1 teorie.<\/p>\n

\n

Projekty a case studies<\/h2>\n

\u2699\ufe0f Unified ML Pipeline<\/h3>\n

N\u00e1vrh end-to-end ML pipeline na platform\u011b Databricks pro bankovn\u00ed instituci. V\u00fdsledek: nasazen\u00ed model\u016f zkr\u00e1ceno z 10 na 2 dny, p\u0159esnost model\u016f +15 %, pipeline \u221240 % rychleji.<\/p>\n

Technologie:<\/strong> Python, PySpark, Databricks, MLflow, Optuna, Docker<\/p>\n

\n
\n Detailn\u00ed case study \u2192<\/a>\n <\/div>\n<\/div>\n

\ud83c\udfe0 PENB Energy Label \u2013 odhad energetick\u00e9 n\u00e1ro\u010dnosti<\/h3>\n

Webov\u00e1 aplikace, kter\u00e1 odhaduje t\u0159\u00eddu energetick\u00e9 n\u00e1ro\u010dnosti bytu z provozn\u00edch dat \u2013 bez nutnosti form\u00e1ln\u00edho auditu. Kombinuje termodynamick\u00fd RC model s meteorologick\u00fdmi daty.<\/p>\n

Dostupn\u00e9 online:<\/strong> penb.michaelprinc.com<\/a><\/p>\n

\n
\n V\u00edce o projektu \u2192<\/a>\n <\/div>\n<\/div>\n

\ud83e\udd16 MCP Prompt Broker<\/h3>\n

Open source sm\u011brovac\u00ed syst\u00e9m pro instruk\u010dn\u00ed profily AI agent\u016f na b\u00e1zi protokolu MCP (Model Context Protocol). Frontier AI engineering projekt pro 2025\u20132026.<\/p>\n

GitHub:<\/strong> github.com\/michaelprinc\/MCP_prompt_broker<\/a><\/p>\n

\n
\n V\u00edce o projektu \u2192<\/a>\n <\/div>\n<\/div>\n
\n
\n Zobrazit cel\u00e9 portfolio \u2192<\/a>\n <\/div>\n<\/div>\n
\n

Portfolio v\u00fdsledk\u016f \u2013 \u010d\u00edsla, kter\u00e1 mluv\u00ed<\/h2>\n\n\n\n\n\n\n\n\n\n\n
Co jsem dos\u00e1hl<\/th>\nV\u00fdsledek<\/th>\n<\/tr>\n<\/thead>\n
Zlep\u0161en\u00ed lift statistiky po aktualizaci produk\u010dn\u00edch model\u016f<\/td>\n+30 %<\/strong><\/td>\n<\/tr>\n
Zkr\u00e1cen\u00ed doby nasazen\u00ed ML model\u016f<\/td>\nz 10 na 2 dny (\u221280 %)<\/strong><\/td>\n<\/tr>\n
Zrychlen\u00ed ML pipeline \u2013 optimalizace datov\u00fdch zdroj\u016f<\/td>\n\u221240 %<\/strong><\/td>\n<\/tr>\n
Zlep\u0161en\u00ed kampa\u0148ov\u00e9 akvizice<\/td>\n+7 %<\/strong><\/td>\n<\/tr>\n
Produk\u010dn\u00edch ML model\u016f v Unified Pipeline<\/strong><\/td>\n~100<\/strong><\/td>\n<\/tr>\n
Let zku\u0161enost\u00ed v DS\/ML<\/strong><\/td>\n7+<\/strong><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
\n

Praktick\u00e9 workshopy o AI \u2013 pro v\u00e1s i va\u0161i firmu<\/h2>\n

V \u00fanoru 2026 prob\u011bhl \u00fasp\u011b\u0161n\u00fd praktick\u00fd workshop o AI v caf\u00e9 Cafedu ve Vinohrady. Workshopy ved\u00edm v mal\u00fdch skupin\u00e1ch (5\u201310 osob) \u2013 obsah spoluvytv\u00e1\u0159\u00edm s \u00fa\u010dastn\u00edky podle jejich skute\u010dn\u00fdch pot\u0159eb.<\/p>\n

T\u00e9mata workshop\u016f:<\/strong><\/p>\n