{"id":622,"date":"2026-09-06T11:06:51","date_gmt":"2026-09-06T14:06:51","guid":{"rendered":"https:\/\/www.linuxpro.com.br\/2026\/09\/data-science-na-vida-real-grupy-sp\/"},"modified":"2026-09-06T11:06:51","modified_gmt":"2026-09-06T14:06:51","slug":"data-science-na-vida-real-grupy-sp","status":"publish","type":"post","link":"https:\/\/www.linuxpro.com.br\/en\/2026\/09\/data-science-na-vida-real-grupy-sp\/","title":{"rendered":"Data Science in real life (GruPy-SP)"},"content":{"rendered":"<p><img decoding=\"async\" alt=\"Sal\u00e3o de meetup Python com o logo oficial das duas cobras\" src=\"https:\/\/www.linuxpro.com.br\/wp-content\/uploads\/2026\/09\/data-science-na-vida-real-grupy-sp-hq.webp\" \/><\/p>\n<p>On October 24, 2015, GruPy-SP (with InteropMIX) heard Fabio Franco Costa talk about data science without a slide promising \u201cin three months you will be a scientist.\u201d His thesis fits in one sentence: <strong>data science is a method for thinking through problems<\/strong>, not a stack. The video is on the LinuxPro channel. Below is the talk's message \u2014 and what still holds true in 2026.<\/p>\n<p><!-- more --><\/p>\n<p><iframe style=\"width:100%;aspect-ratio:16\/9;border:0\" src=\"https:\/\/www.youtube-nocookie.com\/embed\/V6vx6GXMolU\" allowfullscreen loading=\"lazy\" title=\"GruPy-SP | Fabio Franco Costa - Data Science in Real Life\"><\/iframe><\/p>\n<p>Source: <a href=\"https:\/\/www.youtube.com\/watch?v=V6vx6GXMolU\">GruPy-SP | Fabio Franco Costa \u2014 Data Science in Real Life<\/a> (26 min). That same evening: <a href=\"https:\/\/www.youtube.com\/watch?v=eZrboFfBDFw\">Bruno Rocha \u2014 Flask and MongoDB<\/a>.<\/p>\n<div style=\"background:#fff8e1;border-left:4px solid #8b1a1a;padding:.8em 1em;margin:1em 0\"><strong>Context:<\/strong> A talk from 2015. CI&amp;T in Campinas, pandas and scikit-learn at their peak. The method (hypothesis, experiment, statistics, domain knowledge) has not aged. The PPA and Hadoop of that era have. Python 3.14 and this blog's current tutorial: <a href=\"\/en\/2017\/04\/python-do-basico-ao-profissional\/\">Python from Beginner to Professional<\/a>.<\/div>\n<h2>Who spoke, and where<\/h2>\n<p>Fabio Franco Costa, then at <strong>CI&amp;T<\/strong> \u2014 a software company from Campinas that was already a Brazilian multinational, with around two thousand employees and work mentioned for FIFA and Coca-Cola FEMSA. He presents himself as someone who translates data science for sales and the team: he is not the guy who \u201cknows the answer,\u201d but the one who sets up the test. His Twitter at the time: <code data-no-translation=\"\">@costa_2_0<\/code>. Meetup: <a href=\"https:\/\/www.meetup.com\/Grupy-SP\/\">GruPy-SP<\/a>.<\/p>\n<p>The title plays on his own story: data science \u201cin real life\u201d \u2014 including the part where he describes using the same method outside the office. It is not a pandas course. It is the craft.<\/p>\n<h2>The first principle: you do not know the answer<\/h2>\n<p>The barrier, Fabio says, is not the library. It is the attitude of the \u201cdoctor who has an answer for everything.\u201d In academia, you are trained to tackle a problem <em>without<\/em> knowing the function: formulate a hypothesis, measure, fail, formulate again. In business, the opposite is the default \u2014 opinion, rank, \u201cwe have always done it this way.\u201d.<\/p>\n<p>Data science starts when someone admits: <strong>I do not know<\/strong>. Then come the data (almost every company already has tables; an empty starting point is rare), and intuition becomes a question, not a conclusion. He cites the kind of question businesses were already asking in 2015: can you tell whether a customer is pregnant just from her shopping cart? The Target anecdote (USA) was the meme of the time. The point is not American retail \u2014 it is turning a hunch into a testable hypothesis.<\/p>\n<h2>Hypothesis, then experiment<\/h2>\n<p>Having a theory proves nothing. Only the experiment decides. Fabio uses the <strong>black swan<\/strong> in the Popper\/Taleb sense: \u201cevery swan is white\u201d falls apart with a single black swan caught on film. Until you test it, you have a story.<\/p>\n<p>The personal analogy in the talk is an online dating funnel treated as a pipeline: you do not know the DNA of the answer; you build hypotheses (what gets a response, what does not); experiment for a year; measure; discover a side effect (the dinner at home that was not in the sample). The message for the data team is the same: an unexpected result is not a notebook failure \u2014 it is the world.<\/p>\n<p>The third step, which he roughly calls <em>walking<\/em>: if there are no data, you can still design the test. Without a test, statistics cannot save you.<\/p>\n<h2>Statistics are not decoration<\/h2>\n<p>The more tests you run, the greater the chance of a false positive. Fabio mentions the confidence level of 5%: across dozens of experiments, some \u201cwinners\u201d are noise. Without statistics, you are not discussing things with the decision-maker \u2014 you are pushing a chart. The scientist advises; the commander decides. Swapping those roles is the shortcut to a vanity dashboard.<\/p>\n<p>The trio he advocates: <strong>statistics + business knowledge + research method<\/strong>. Tools alone (the \u201cmonkey test\u201d) do not get to the problem. Domain knowledge alone without testing becomes trained guesswork. Testing alone without domain knowledge optimizes the wrong metric.<\/p>\n<h2>The simple scoring example<\/h2>\n<p>At the end, he points to a classic pediatric test \u2014 what the audience recognizes as <strong>Apgar<\/strong> (Virginia Apgar, 1952): a few objective questions about the newborn, a score from 0 to 10, rather than just the on-duty doctor's \u201cclinical eye.\u201d The literature of the time already showed that a simple checklist, applied immediately, reduces first-day mortality compared with a specialist's unstructured impression. Fabio's message: a humble model + data in the right place often performs better than a well-trained human on autopilot.<\/p>\n<p>It is not \u201cAI replaces the doctor.\u201d It is this: the scientific method fits any initiative, and the best place to apply it is where there is no legacy yet \u2014 a new product, a new process, an internal startup.<\/p>\n<h2>A data-driven organization<\/h2>\n<p>His \u201cpersonal crusade\u201d in 2015: helping companies make <em>some<\/em> decisions with humility and evidence, not just intuition. Data-driven does not mean BI on the wall. It means reducing the number of choices made in the dark.<\/p>\n<p>CI&amp;T was launching a product based on data science \/ machine learning in the following cycle. The name on the box matters less than the habit: a written hypothesis, a designed experiment, statistics in the middle, domain knowledge at the end.<\/p>\n<h2>What to do with this on Linux in 2026<\/h2>\n<p>The setting was Python \u2014 GruPy. The local path remains the same, with newer versions:<\/p>\n<pre data-no-translation=\"\"><code class=\"language-bash\" data-no-translation=\"\">python3 -m venv .venv\nsource .venv\/bin\/activate\npython -m pip install -U pip pandas scikit-learn matplotlib\n<\/code><\/pre>\n<p>Tutorial and venv: <a href=\"\/en\/2017\/04\/python-do-basico-ao-profissional\/\">Python from Beginner to Professional<\/a>. The language overview: <a href=\"\/en\/2017\/05\/linguagens-de-programacao\/\">The 7 main programming languages<\/a>.<\/p>\n<p>Pandas and sklearn still handle everyday tabular work. What has changed is the number of ready-made models (and people who skip testing). The 2015 talk is the antidote: <strong>I do not know \u2192 hypothesis \u2192 experiment \u2192 a number that stands up to discussion<\/strong>.<\/p>\n<h2>Wrap-up<\/h2>\n<p>A GruPy Saturday, a guy from CI&amp;T, a method academia already teaches and business forgets. Watch the video. Then open a dataset of your own and write the hypothesis <em>before<\/em> the chart.<\/p>\n<p>GruPy-SP: <a href=\"https:\/\/www.meetup.com\/Grupy-SP\/\">meetup.com\/Grupy-SP<\/a> \u00b7 channel: <a href=\"https:\/\/www.youtube.com\/linuxpro\">youtube.com\/linuxpro<\/a>.<\/p>","protected":false},"excerpt":{"rendered":"<p>On October 24, 2015, GruPy-SP (with InteropMIX) heard Fabio Franco Costa discuss data science without slides promising \u201cyou will be a scientist in three months.\u201d The argument fits in one sentence: data science is a method for thinking about problems, not a stack. The video is on the LinuxPro channel. Below is the talk\u2019s message \u2014 \u2026 <a title=\"Data Science in real life (GruPy-SP)\" class=\"read-more\" href=\"https:\/\/www.linuxpro.com.br\/en\/2026\/09\/data-science-na-vida-real-grupy-sp\/\" aria-label=\"Read more about Data Science na vida real (GruPy-SP)\">Read more<\/a><\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[45,80],"tags":[283,281,282,284,81],"class_list":["post-622","post","type-post","status-publish","format-standard","hentry","category-desenv","category-python","tag-ciencia-de-dados","tag-data-science","tag-grupy","tag-meetup","tag-python"],"_links":{"self":[{"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/posts\/622","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/comments?post=622"}],"version-history":[{"count":1,"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/posts\/622\/revisions"}],"predecessor-version":[{"id":623,"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/posts\/622\/revisions\/623"}],"wp:attachment":[{"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/media?parent=622"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/categories?post=622"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.linuxpro.com.br\/en\/wp-json\/wp\/v2\/tags?post=622"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}