diff --git a/README.md b/README.md
index c7f1103..4fda2c5 100644
--- a/README.md
+++ b/README.md
@@ -2,6 +2,11 @@
 ## 📊 simulate_calibrate_sir() and 🔧 calibrate_sir()
 
 ### ✍️ Authors: George Vega Yon, Sima NJF
+# 🌍 Introduction
+
+Predicting the trajectory of infectious diseases lies at the heart of public health preparedness. With the epiworldRcalibrate package, you can effortlessly simulate and calibrate SIR (Susceptible-Infected-Recovered) epidemic models, blending the power of R, TensorFlow, and Keras. You can swiftly bridge the gap between theory and practice by generating realistic epidemic scenarios and training advanced Convolutional Neural Networks (CNNs) to identify key parameters—like prevalence, contact rate, transmission probability, and recovery precision.
+
+The 🚀 simulate_calibrate_sir() function empowers you to produce synthetic incidence data and refine your model’s parameters through deep learning. Complementing this, 🔧 calibrate_sir() takes observed data, returning carefully tuned parameters ready for further analysis. Together, these tools streamline the modeling-to-inference pipeline, guiding you toward more informed, data-driven decisions in epidemiological research and public health policy.
 
 ### 🚀 simulate_calibrate_sir()
 
@@ -31,42 +36,6 @@ epochs <- 2
 verbose <- 2
 ```
 ```{r}
-simulate_calibrate_sir <- function(N, n, ndays, ncores, epochs, verbose) {
-  # ⚙️ Generate Theta and Seeds
-  theta <- generate_theta(N, n)
-  seeds <- sample.int(.Machine$integer.max, N, TRUE)
-
-  # 🧪 Run Simulations
-  matrices <- run_simulations(N, n, ndays, ncores, theta, seeds)
-
-  # 🔍 Filter Non-Null Data
-  filtered_data <- filter_non_null(matrices, theta)
-  matrices <- filtered_data$matrices
-  theta <- filtered_data$theta
-  N <- filtered_data$N
-
-  # 📊 Prepare Data for TensorFlow
-  arrays_1d <- prepare_data_for_tensorflow(matrices, N)
-  theta2 <- as.data.table(copy(theta))
-  theta2$crate <- plogis(theta2$crate / 10)
-
-  # 🔀 Split Data into Training and Testing Sets
-  data_split <- split_data(arrays_1d, theta2, N)
-  train <- data_split$train
-  test <- data_split$test
-
-  # 🏗️ Build and Train the CNN Model
-  model <- build_cnn_model(dim(arrays_1d)[-1], ncol(theta))
-  train_model(model, train, epochs = epochs, verbose = verbose)
-
-  # 📈 Evaluate the Model
-  eval_results <- evaluate_model(model, test, theta)
-  pred <- eval_results$pred
-  MAEs <- eval_results$MAEs
-  plot_results(pred, test, theta, MAEs, N, floor(N * 0.7))
-  return(list(pred = pred, MAEs = MAEs))
-}
-
 # ▶️ Call the function
 simulate_calibrate_sir(N, n, ndays, ncores, epochs, verbose)
 ```