diff --git a/Dockerfile b/Dockerfile
new file mode 100644
index 0000000..fa0a6a9
--- /dev/null
+++ b/Dockerfile
@@ -0,0 +1,23 @@
+# app/Dockerfile
+
+FROM python:3.9-slim
+
+# Copy Code
+WORKDIR /wise_pizza
+
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    software-properties-common \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+COPY . .
+
+RUN pip3 install -r docker_requirements.txt
+
+EXPOSE 8501
+
+HEALTHCHECK CMD curl --fail http://localhost:8501/_stcore/health
+
+ENTRYPOINT ["streamlit", "run", "streamlit_app.py", "--server.port=8501", "--server.address=0.0.0.0"]
\ No newline at end of file
diff --git a/README.md b/README.md
index 4623ecd..a4d829c 100644
--- a/README.md
+++ b/README.md
@@ -34,6 +34,8 @@ Sometimes, rather than explaining the change in totals from one period to the ne
     - [Comparison between two datasets](#understanding-differences-in-two-time-periods-or-two-dataframes)
 - [Installation](#installation)
 - [Quick Start](#quick-start)
+- [Streamlit app](#streamlit-app)
+    - [Docker container](#docker-container)
 - [For Developers](#for-developers)
  - [Tests](#testing)
 
@@ -134,10 +136,31 @@ sf.segments
 ```
 Please see the full example [here](https://github.com/transferwise/wise-pizza/blob/main/notebooks/Finding%20interesting%20segments.ipynb)
 
-## For Developers
+## Streamlit app
+In the root directory of this repository, there is a Streamlit app. This is an interface that allows you to upload your own files and run analyses, as you saw in the Jupyter Notebook provided as an example.
+
+To run this, you need to:
+1. Create a virtual environment (e.g. using pyenv)
+2. Activate the virtual environment
+3. Run `pip -r requirements.txt` before running, to install necessary dependencies
+4. Run `streamlit run streamlit_app.py` to run Streamlit
+
+### Docker container
 
+We created a Docker container that makes it easier to deploy this solution elsewhere
 
+You need to first: Create the Docker image
+Create the Docker image
 
+```Python
+docker build -t streamlit .      
+```
+And then simply run the image
+
+```Python
+docker run -p 8501:8501 streamlit    
+```
+## For Developers
 
 ### Testing
 We use [PyTest](https://docs.pytest.org/) for testing. If you want to contribute code, make sure that the tests in tests/ run without errors.
diff --git a/docker_requirements.txt b/docker_requirements.txt
new file mode 100644
index 0000000..0f668d9
--- /dev/null
+++ b/docker_requirements.txt
@@ -0,0 +1,12 @@
+ipython==8.14.0
+kaleido==0.2.1
+numpy==1.24.0
+pandas==2.0.2
+pytest==7.4.3
+plotly==5.15.0
+scikit_learn==1.3.2
+scipy==1.8.0
+tqdm==4.66.1
+cloudpickle==3.0.0
+pivottablejs==0.9.0
+streamlit==1.28.0
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
index 528bed9..18b5aba 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -9,3 +9,4 @@ scipy>=1.8.0
 tqdm
 cloudpickle
 pivottablejs
+streamlit==1.28.0
\ No newline at end of file
diff --git a/streamlit_app.py b/streamlit_app.py
new file mode 100644
index 0000000..a2e2f13
--- /dev/null
+++ b/streamlit_app.py
@@ -0,0 +1,290 @@
+"""An example of Streamlit leveraging Wise pizza."""
+
+import streamlit as st
+import os, sys
+import pandas as pd
+from wise_pizza.explain import explain_levels, explain_changes_in_totals, explain_changes_in_average
+
+
+# False if you want nice interactive plots
+# True if you want static plots (Doesn't work on all platforms yet)
+plot_is_static = False
+
+
+# SETTING PAGE CONFIG TO WIDE MODE AND ADDING A TITLE AND FAVICON
+st.set_page_config(layout="wide", page_title="Wise Pizza", page_icon=":pizza:")
+
+st.title('Wise Pizza powered by Streamlit')
+st.text('Only categorical columns are accepted, bucket the numeric ones if you wanna use those')
+
+# upload the file from the computer
+def load_data_upload():
+    uploaded_file = st.file_uploader("Choose a file")
+    if not uploaded_file:
+        st.warning('Please input a dataset.')
+        st.stop()
+    st.success('Dataset inputted.')
+    data = pd.read_csv(uploaded_file)
+    return data
+
+on = st.toggle('Use sample data from Github')
+url_data = (r'https://raw.githubusercontent.com/transferwise/wise-pizza/main/data/synth_data.csv')
+
+# select the datasource, either local or from github
+if on:
+    st.write(f'Downloading data from {url_data}!')
+    df = pd.read_csv(url_data)
+else:
+    df=load_data_upload()
+
+# show dataset preview
+st.text('Table preview')
+st.table(df.head(10))
+
+# ask the user for relevant dimensions
+dims = st.multiselect(
+    "Select the dimensions you want to include in the analysis",
+    df.select_dtypes(exclude=['number']).columns.tolist()
+    )
+
+# ask the user via streamlit to select if they want to run a comparison between subgroups or not
+flag_comparison = st.toggle('I want to run a comparison between two subgroups in my data')
+
+# return columns that are candidate for comparison
+def flag_columns(df):
+    binary_columns = df.columns[df.nunique() == 2].tolist()
+    if not binary_columns:
+        st.warning('No column in the dataset is binary, no comparison can be carried out')
+        st.stop()
+    return binary_columns
+
+if flag_comparison:
+    #calculate binary columns
+    binary_columns=flag_columns(df)
+
+    # ask the user via streamlit for the flag column
+    flag_column = st.selectbox(
+    "What is the flag column of your dataset that defines the two subgroups?",
+    binary_columns,
+    index=None,
+    placeholder="Select flag column column",
+    )
+    st.write('You selected:', flag_column)
+
+    # wait until flag column is added
+    if not flag_column:
+            st.stop()
+
+    # calculate unique flags in the dataset
+    flags = sorted(df[flag_column].unique())  
+
+    # allow users to define what's "old" and "new" in the comparison
+    flags_option = st.selectbox(
+        'Which value in your flag column belongs to group A?',
+        (flags))
+
+
+# ask the user via streamlit for the target column
+totals = st.selectbox(
+   "Name of column that contains totals per segment (e.g. GMV/revenue)",
+   # display only numerical columns
+   df.select_dtypes(include=['number']).columns.tolist(),
+   index=None,
+   placeholder="Select target column",
+)
+st.write('You selected:', totals)
+
+# ask the user via streamlit for the size column
+size = st.selectbox(
+   "Name of column containing segment size (e.g. number of users/number of transactions)",
+   # display only numerical columns
+   df.select_dtypes(include=['number']).columns.tolist(),
+   index=None,
+   placeholder="Select volume column",
+)
+st.write('You selected:', size)
+
+st.subheader('Finding the juiciest slices', divider='rainbow')
+st.text('This section does not compare groups, but rather checks which features have the most impact in the target column you selected.')
+
+if not totals or not size or not dims:
+        st.warning('Please input all fields above.')
+        st.stop()
+
+## finding juiciest slices
+sf = explain_levels(
+    df=df,
+    dims=dims,
+    total_name=totals,
+    size_name=size,
+    max_depth=2,
+    min_segments=20,
+    solver="lasso",
+    return_fig=True
+)
+
+# storing the plot in a variable
+plot_sf=sf.plot(width=500, height=500)
+# exposing the plot via streamlit
+st.plotly_chart(plot_sf, use_container_width=True)
+
+if flag_comparison:
+    st.subheader('Analysing differences', divider='rainbow')
+    st.text('This section does compare the two groups defined by the flag. Old total is the group A you selected in the dropdown')
+
+    # creating the dataframes for the comparison calculations
+    data = df[df[flag_column] != flags_option]  # take the group to compare to
+    pre_data = df[df[flag_column] == flags_option]  # take the group to be compared
+
+    # define the relevant dimensions for the comparison feature
+    # for this, the flag column is to be excluded
+    comparison_dimensions = list(filter(lambda x: x != flag_column, dims))
+
+    ## running explain calculations
+    sf1 = explain_changes_in_totals(
+        df1=pre_data,
+        df2=data,
+        dims=comparison_dimensions,
+        total_name=totals,
+        size_name=size,
+        max_depth=2,
+        min_segments=20,
+        how="totals",
+        solver="lasso",
+        return_fig=True
+    )
+    # specifying a two column layout on streamlit
+    col1, col2 = st.columns(2)
+    # storing the plots in variables
+    # exposing the plots via streamlit
+    with col1:
+        plot_sf1=sf1.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+        st.plotly_chart(plot_sf1, use_container_width=True)
+
+    with col2:
+        plot_sf2=sf1.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+        st.plotly_chart(plot_sf2, use_container_width=True)
+
+    st.subheader('Decomposing differences', divider='rainbow')
+    st.text('`split_fits` to separately decompose contribution of size changes and average changes')
+
+    ## running explain calculations
+    sf2 = explain_changes_in_totals(
+        df1=pre_data,
+        df2=data,
+        dims=comparison_dimensions,
+        total_name=totals,
+        size_name=size,
+        max_depth=1,
+        min_segments=10,
+        how="split_fits",
+        solver="lasso",
+        return_fig=True
+    )
+    # storing the plot in a variable
+    # exposing the plot via streamlit
+    plot_sf=sf2.plot(width=500, height=500)
+    st.plotly_chart(plot_sf, use_container_width=True)
+
+    st.text('`extra_dim` to treat size vs average change contribution as an additional dimension')
+
+    ## running explain calculations
+    sf3 = explain_changes_in_totals(
+        df1=pre_data,
+        df2=data,
+        dims=comparison_dimensions,
+        total_name=totals,
+        size_name=size,
+        max_depth=2,
+        min_segments=20,
+        how="extra_dim",
+        solver="lasso",
+        return_fig=True
+    )
+    # specifying a two column layout on streamlit
+    col1, col2 = st.columns(2)
+
+    # storing the plots in variables
+    # exposing the plots via streamlit
+    with col1:
+        plot_sf1=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+        st.plotly_chart(plot_sf1, use_container_width=True)
+
+    with col2:
+        plot_sf2=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+        st.plotly_chart(plot_sf2, use_container_width=True)
+
+    st.text('`force_dim` like extra_dim, but each segment must contain a Change_from constraint')
+
+    ## running explain calculations
+    sf3 = explain_changes_in_totals(
+        df1=pre_data,
+        df2=data,
+        dims=comparison_dimensions,
+        total_name=totals,
+        size_name=size,
+        max_depth=2,
+        min_segments=15,
+        how="force_dim",
+        solver="lasso",
+        return_fig=True
+    )
+    # specifying a two column layout on streamlit
+    col1, col2 = st.columns(2)
+
+    # storing the plots in variables
+    # exposing the plots via streamlit
+    with col1:
+        plot_sf1=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+        st.plotly_chart(plot_sf1, use_container_width=True)
+
+    with col2:
+        plot_sf2=sf3.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+        st.plotly_chart(plot_sf2, use_container_width=True)
+
+    st.subheader('Explaining changes in average', divider='rainbow')
+
+    ## running explain calculations
+    sf4 = explain_changes_in_average(
+        df1=pre_data,
+        df2=data,
+        dims=comparison_dimensions,
+        total_name=totals,
+        size_name=size,
+        max_depth=2,
+        min_segments=20,
+        how="totals",
+        solver="lasso",
+        return_fig=True
+    )
+    # specifying a two column layout on streamlit
+    col1, col2 = st.columns(2)
+
+    # storing the plots in variables
+    # exposing the plots via streamlit
+    with col1:
+        plot_sf1=sf4.plot(width=500, height=500, plot_is_static=plot_is_static)[0]
+        st.plotly_chart(plot_sf1, use_container_width=True)
+
+    with col2:
+        plot_sf2=sf4.plot(width=500, height=500, plot_is_static=plot_is_static)[1]
+        st.plotly_chart(plot_sf2, use_container_width=True)
+
+    ## running explain calculations
+    sf6 = explain_changes_in_average(
+        df1=pre_data,
+        df2=data,
+        dims=comparison_dimensions,
+        total_name=totals,
+        size_name=size,
+        max_depth=2,
+        min_segments=20,
+        how="split_fits",
+        solver="lasso",
+        return_fig=True
+    )
+
+    # storing the plot in a variable
+    plot_sf=sf6.plot(width=500, height=500)
+    # exposing the plot via streamlit
+    st.plotly_chart(plot_sf, use_container_width=True)
\ No newline at end of file
diff --git a/wise_pizza/explain.py b/wise_pizza/explain.py
index 4c2c95b..5834b2b 100644
--- a/wise_pizza/explain.py
+++ b/wise_pizza/explain.py
@@ -26,6 +26,7 @@ def explain_changes_in_average(
     force_add_up: bool = False,
     constrain_signs: bool = True,
     verbose: int = 0,
+    return_fig: bool = False
 ):
     """
     Find segments most useful in explaining the difference between the averages of the two datasets
@@ -48,6 +49,7 @@ def explain_changes_in_average(
     @param constrain_signs: Whether to constrain weights of segments to have the same
     sign as naive segment averages
     @param verbose: If set to a truish value, lots of debug info is printed to console
+    @param return_fig: If set to true, plot returns the figure object, otherwise shows the figures
     @return: A fitted object
     """
     df1 = df1.copy()
@@ -86,6 +88,7 @@ def explain_changes_in_average(
         force_add_up=force_add_up,
         constrain_signs=constrain_signs,
         verbose=verbose,
+        return_fig=return_fig
     )
 
     if hasattr(sf, "pre_total"):
@@ -119,6 +122,7 @@ def explain_changes_in_totals(
     force_add_up: bool = False,
     constrain_signs: bool = True,
     verbose: int = 0,
+    return_fig: bool = False
 ):
     """
     Find segments most useful in explaining the difference between the totals of the two datasets
@@ -141,6 +145,7 @@ def explain_changes_in_totals(
     @param constrain_signs: Whether to constrain weights of segments to have the same
     sign as naive segment averages
     @param verbose: If set to a truish value, lots of debug info is printed to console
+    @param return_fig: If set to true, plot returns the figure object, otherwise shows the figures
     @return: A fitted object
     """
 
@@ -181,6 +186,7 @@ def explain_changes_in_totals(
             force_add_up=force_add_up,
             constrain_signs=constrain_signs,
             verbose=verbose,
+            return_fig=return_fig
         )
 
         sf_avg = explain_levels(
@@ -195,6 +201,7 @@ def explain_changes_in_totals(
             force_add_up=force_add_up,
             constrain_signs=constrain_signs,
             verbose=verbose,
+            return_fig=return_fig
         )
 
         sf_size.final_size = final_size
@@ -207,6 +214,7 @@ def explain_changes_in_totals(
                 plot_is_static=plot_is_static,
                 width=width,
                 height=height,
+                return_fig=return_fig
             )
         )
         return sp
@@ -233,7 +241,7 @@ def explain_changes_in_totals(
         sf.post_total = df2[total_name].sum()
 
         sf.plot = lambda plot_is_static=False, width=1000, height=1000: plot_waterfall(
-            sf, plot_is_static=plot_is_static, width=width, height=height
+            sf, plot_is_static=plot_is_static, width=width, height=height, return_fig=return_fig
         )
         sf.task = "changes in totals"
         return sf
@@ -252,6 +260,7 @@ def explain_levels(
     verbose=0,
     force_add_up: bool = False,
     constrain_signs: bool = True,
+    return_fig: bool = False
 ):
     """
     Find segments whose average is most different from the global one
@@ -267,6 +276,7 @@ def explain_levels(
     @param verbose: If set to a truish value, lots of debug info is printed to console
     @param force_add_up: Force the contributions of chosen segments to add up to zero
     @param constrain_signs: Whether to constrain weights of segments to have the same
+    @param return_fig: If set to true, plot returns the figure object, otherwise shows the figures
     sign as naive segment averages
     @return: A fitted object
     """
@@ -304,7 +314,7 @@ def explain_levels(
         s["total"] += average * s["seg_size"]
     # print(average)
     sf.reg.intercept_ = average
-    sf.plot = lambda plot_is_static=False, width=2000, height=500, return_fig=False: plot_segments(
+    sf.plot = lambda plot_is_static=False, width=2000, height=500, return_fig=return_fig: plot_segments(
         sf, plot_is_static=plot_is_static, width=width, height=height, return_fig=return_fig
     )
     sf.task = "levels"
diff --git a/wise_pizza/plotting.py b/wise_pizza/plotting.py
index 7b4ce04..86db39c 100644
--- a/wise_pizza/plotting.py
+++ b/wise_pizza/plotting.py
@@ -20,6 +20,7 @@ def plot_split_segments(
     plot_is_static: bool = False,
     width: int = 2000,
     height: int = 500,
+    return_fig: bool = False
 ):
     """
     Plot split segments for explain_changes: split_fits
@@ -123,7 +124,10 @@ def plot_split_segments(
             width=width + len(size_data.index) * 30,
         )
     else:
-        fig.show()
+        if return_fig:
+            return fig
+        else:
+            fig.show()
 
 
 def plot_segments(
@@ -274,7 +278,11 @@ def waterfall_layout_args(sf: SliceFinder, width: int = 1000, height: int = 1000
 
 
 def plot_waterfall(
-    sf: SliceFinder, plot_is_static: bool = False, width: int = 1000, height: int = 1000
+    sf: SliceFinder, 
+    plot_is_static: bool = False, 
+    width: int = 1000, 
+    height: int = 1000,
+    return_fig: bool = False
 ):
     """
     Plot waterfall and Bar for explain_changes
@@ -312,14 +320,17 @@ def plot_waterfall(
         **waterfall_layout_args(sf, width, height)
     )
 
-    if plot_is_static:
-        # Convert the figure to a static image
-        image_bytes = to_image(fig, format="png", scale=2)
-        image_bytes2 = to_image(fig2, format="png", scale=2)
-
-        # Display the static image in the Jupyter notebook
-        display(Image(image_bytes, width=width, height=height))
-        display(Image(image_bytes2, width=width, height=height))
+    if return_fig:
+            return [fig, fig2]
     else:
-        fig.show()
-        fig2.show()
+        if plot_is_static:
+            # Convert the figure to a static image
+            image_bytes = to_image(fig, format="png", scale=2)
+            image_bytes2 = to_image(fig2, format="png", scale=2)
+
+            # Display the static image in the Jupyter notebook
+            display(Image(image_bytes, width=width, height=height))
+            display(Image(image_bytes2, width=width, height=height))
+        else:
+            fig.show()
+            fig2.show()
\ No newline at end of file