diff --git a/src/core/algorithms/algo_factory.cpp b/src/core/algorithms/algo_factory.cpp
index ac4439a087..9b990181f3 100644
--- a/src/core/algorithms/algo_factory.cpp
+++ b/src/core/algorithms/algo_factory.cpp
@@ -9,6 +9,7 @@
 #include "algorithms/pipelines/typo_miner/typo_miner.h"
 #include "config/names.h"
 #include "tabular_data/input_tables_type.h"
+#include "util/separator_validator.h"
 
 namespace algos {
 
@@ -47,10 +48,12 @@ void LoadAlgorithm(Algorithm& algorithm, StdParamsMap const& options) {
         using namespace config::names;
         namespace fs = std::filesystem;
         if (option_name == kTable && options.find(std::string{kTable}) == options.end()) {
-            config::InputTable parser =
+            auto csv_parser =
                     std::make_shared<CSVParser>(GetOptionValue<fs::path>(options, kCsvPath),
                                                 GetOptionValue<char>(options, kSeparator),
                                                 GetOptionValue<bool>(options, kHasHeader));
+            csv_parser->ValidateSeparator();
+            config::InputTable parser = csv_parser;
             return boost::any{parser};
         } else if (option_name == kTables && options.find(std::string{kTables}) == options.end()) {
             auto paths = GetOptionValue<std::vector<fs::path>>(options, kCsvPaths);
diff --git a/src/core/parser/csv_parser/csv_parser.cpp b/src/core/parser/csv_parser/csv_parser.cpp
index 6ee949e0bf..360a53803b 100644
--- a/src/core/parser/csv_parser/csv_parser.cpp
+++ b/src/core/parser/csv_parser/csv_parser.cpp
@@ -1,14 +1,17 @@
 #include "csv_parser.h"
 
+#include <algorithm>
 #include <cassert>
 #include <filesystem>
 #include <fstream>
 #include <string>
+#include <unordered_map>
 #include <utility>
 #include <vector>
 
 #include <boost/algorithm/string.hpp>
 #include <boost/tokenizer.hpp>
+#include <easylogging++.h>
 
 inline std::string& CSVParser::rtrim(std::string& s) {
     boost::trim_right(s);
@@ -137,3 +140,90 @@ std::vector<std::string> CSVParser::GetNextRow() {
 
     return result;
 }
+
+std::optional<char> CSVParser::DeduceSeparator() {
+    /* Calculate statistics including the header row */
+    bool has_header_copy = has_header_;
+    has_header_ = false;
+    Reset();
+    has_header_ = has_header_copy;
+
+    std::unordered_map<char, unsigned> letter_count;
+    if (has_next_) {
+        for (char c : next_line_) {
+            letter_count[c]++;
+        }
+    }
+
+    std::unordered_map<char, unsigned> next_letter_count;
+    while (has_next_) {
+        GetNextIfHas();
+        next_letter_count.clear();
+        for (char c : next_line_) {
+            next_letter_count[c]++;
+        }
+        for (auto letter : letter_count) {
+            if (letter.second != next_letter_count[letter.first]) {
+                letter_count[letter.first] = 0;
+            }
+        }
+    }
+
+    for (auto letter : letter_count) {
+        if (letter.second != 0) {
+            Reset();
+            return letter.first;
+        }
+    }
+    Reset();
+
+    return std::nullopt;
+}
+
+bool CSVParser::CheckSeparator() {
+    /* Calculate statistics including the header row */
+    bool has_header_copy = has_header_;
+    has_header_ = false;
+    Reset();
+    has_header_ = has_header_copy;
+
+    unsigned sep_count = 0;
+    if (has_next_) {
+        sep_count = std::count(next_line_.begin(), next_line_.end(), separator_);
+    }
+
+    while (has_next_) {
+        GetNextIfHas();
+        if (sep_count != std::count(next_line_.begin(), next_line_.end(), separator_)) {
+            Reset();
+            return false;
+        }
+    }
+    Reset();
+
+    return true;
+}
+
+std::optional<char> CSVParser::ValidateSeparator() {
+    std::optional<char> possible_separator = DeduceSeparator();
+
+    if (CheckSeparator()) {
+        if (possible_separator == std::nullopt || separator_ == possible_separator ||
+            GetNumberOfColumns() != 1) {
+            return separator_;
+        }
+        LOG(WARNING) << "Inserted separator seems to be wrong";
+        LOG(WARNING) << "Possible separator for the table is: \'" << possible_separator.value()
+                     << "\'";
+        return possible_separator;
+    }
+
+    LOG(WARNING) << "Inserted separator seems to be wrong";
+    if (possible_separator != std::nullopt) {
+        LOG(WARNING) << "Possible separator for the table is: \'" << possible_separator.value()
+                     << "\'";
+        return possible_separator;
+    }
+
+    return std::nullopt;
+}
\ No newline at end of file
diff --git a/src/core/parser/csv_parser/csv_parser.h b/src/core/parser/csv_parser/csv_parser.h
index 2339629e3b..96ffaab5f8 100644
--- a/src/core/parser/csv_parser/csv_parser.h
+++ b/src/core/parser/csv_parser/csv_parser.h
@@ -7,6 +7,7 @@
 
 #include <filesystem>
 #include <fstream>
+#include <optional>
 #include <string>
 #include <vector>
 
@@ -30,6 +31,8 @@ class CSVParser : public model::IDatasetStream {
     std::vector<std::string> ParseString(std::string const& s) const;
     void GetNextIfHas();
     void SkipLine();
+    std::optional<char> DeduceSeparator();
+    bool CheckSeparator();
 
     inline static std::string& rtrim(std::string& s);
 
@@ -42,6 +45,8 @@ class CSVParser : public model::IDatasetStream {
     std::string GetUnparsedLine(unsigned long long const line_index);
     std::vector<std::string> ParseLine(unsigned long long const line_index);
 
+    std::optional<char> ValidateSeparator();
+
     bool HasNextRow() const override {
         return has_next_;
     }
diff --git a/src/core/util/separator_validator.cpp b/src/core/util/separator_validator.cpp
new file mode 100644
index 0000000000..d7e628d204
--- /dev/null
+++ b/src/core/util/separator_validator.cpp
@@ -0,0 +1,12 @@
+#include "separator_validator.h"
+
+#include <easylogging++.h>
+
+namespace util {
+
+std::optional<char> ValidateSeparator(std::filesystem::path const& path, char separator) {
+    auto parser = std::make_unique<CSVParser>(path, separator, false);
+    return parser->ValidateSeparator();
+}
+
+}  // namespace util
\ No newline at end of file
diff --git a/src/core/util/separator_validator.h b/src/core/util/separator_validator.h
new file mode 100644
index 0000000000..2196e750ca
--- /dev/null
+++ b/src/core/util/separator_validator.h
@@ -0,0 +1,12 @@
+#pragma once
+
+#include <filesystem>
+#include <optional>
+
+#include "parser/csv_parser/csv_parser.h"
+
+namespace util {
+
+std::optional<char> ValidateSeparator(std::filesystem::path const& path, char separator);
+
+}  // namespace util
\ No newline at end of file
diff --git a/src/python_bindings/py_algorithm.cpp b/src/python_bindings/py_algorithm.cpp
index 575eeff39a..2f9099f504 100644
--- a/src/python_bindings/py_algorithm.cpp
+++ b/src/python_bindings/py_algorithm.cpp
@@ -14,6 +14,7 @@
 #include "get_py_type.h"
 #include "parser/csv_parser/csv_parser.h"
 #include "py_to_any.h"
+#include "util/separator_validator.h"
 
 namespace python_bindings {
 
@@ -80,7 +81,9 @@ void PyAlgorithmBase::LoadProvidedData(pybind11::kwargs const& kwargs, InputTabl
 
 void PyAlgorithmBase::LoadData(std::string_view path, char separator, bool has_header,
                                py::kwargs const& kwargs) {
-    LoadProvidedData(kwargs, std::make_shared<CSVParser>(path, separator, has_header));
+    auto parser = std::make_shared<CSVParser>(path, separator, has_header);
+    parser->ValidateSeparator();
+    LoadProvidedData(kwargs, parser);
 }
 
 void PyAlgorithmBase::LoadData(py::handle dataframe, std::string name, py::kwargs const& kwargs) {