diff --git a/experiments/lazy_load.md b/experiments/lazy_load.md
index 43009c7f30..d7590be803 100644
--- a/experiments/lazy_load.md
+++ b/experiments/lazy_load.md
@@ -1,8 +1,8 @@
 # Lazy load experiment results
 
-Branch: sb-ta/lazy-load-experiment
-Date: February 10th, 2026
-Owner(s): Stefanni Brasil and Thiago Araujo
+- Branch: sb-ta/lazy-load-experiment
+- Date: February 10th, 2026
+- Owner(s): Stefanni Brasil and Thiago Araujo
 
 ## Impact
 
@@ -22,6 +22,7 @@ To prevent other generators from erroring out due to namespace clashing, some ge
 - no additional dependencies needed
 - code is extremely faster
 - we can enable this as an opt-in configuration
+- allows for users to create external generators in a transparent way
 
 ## Results
 
@@ -29,8 +30,8 @@ Machine specs: Apple M1 Pro 16GB memory on MacOS Sequoia 15.7.3..
 
 profiler:
 
-[LAZY_LOAD=1 bundle exec vernier run -- ruby -e "require 'faker'"](https://share.firefox.dev/46biNAs)
-[bundle exec vernier run -- ruby -e "require 'faker'"](https://share.firefox.dev/4ams9vM)
+- [LAZY_LOAD=1 bundle exec vernier run -- ruby -e "require 'faker'"](https://share.firefox.dev/46biNAs)
+- [bundle exec vernier run -- ruby -e "require 'faker'"](https://share.firefox.dev/4ams9vM)
 
 benchmark:
 
diff --git a/experiments/zeitwerk.md b/experiments/zeitwerk.md
new file mode 100644
index 0000000000..54b0d8d408
--- /dev/null
+++ b/experiments/zeitwerk.md
@@ -0,0 +1,59 @@
+# Autoload with zeitwerk experiment results
+
+- Branch: [sb-autoload-zeitwerk-experiment-3207](https://github.com/faker-ruby/faker/compare/sb-autoload-zeitwerk-experiment-3207?expand=1)
+- Date: February 19th, 2026
+- Owner(s): Stefanni Brasil and Thiago Araujo
+
+## Impact
+
+We want to compare improving faker's performance by lazy loading the generators or autoloading with Zeitwerk. This document shows the benchmarks, and other changes needed to configure Zeitwerk.
+
+With the experiments documented, we can assess the pros and cons of maintainability, fewer breaking changes, easier adoption factors as guiding points for choosing the strategy we will move forward.
+
+### Changes needed
+
+- load and require generators in the correct order
+
+Similarly to the [lazy load experiment](./lazy_load.md), it would be required to load `faker/music` and `faker/internet` first, before the nested namespaces such as `Faker::Music::BossaNova`, as they inherit from class `Music`, for example.
+
+- added a runtime dependency for the library
+
+- Zeitwerk 2.7 requires Ruby >= 3.2. This isn't a deal breaker because we will remove EOL Ruby 3.1 soon, but it would require releasing that version separately.
+
+- our other goal besides improving performance, is allowing users to create their own Faker generators. Zeitwerk scans the file systems to setup the autoloads, so it would not setup autoloads for these external generators.
+
+#### File location changes
+
+To prevent other generators from erroring out due to namespace clashing, some generators have to be moved around (ex. `Faker::Quote` was moved from `/faker/quotes/quote` to `faker/default/quote`). Users can still use the generators as before, their namespaces didn't change.
+
+### Benefits
+
+- less code changes than lazy loading, and lots of customization options available (i.e., eager loading)
+- code is extremely faster, but a bit slower than lazy loading
+- we can enable this as an opt-in configuration
+
+## Results
+
+Machine specs: Apple M1 Pro 16GB memory on MacOS Sequoia 15.7.3.
+
+profiler:
+
+- [AUTOLOAD=1 bundle exec vernier run -- ruby -e "require 'faker'"](https://share.firefox.dev/4aAJJee)
+- [bundle exec vernier run -- ruby -e "require 'faker'"](https://share.firefox.dev/4bWViih)
+
+benchmark:
+
+```sh
+benchmark % ruby load.rb
+ruby 3.3.10 (2025-10-23 revision 343ea05002) [arm64-darwin24]
+Warming up --------------------------------------
+             require     1.000 i/100ms
+            autoload     1.000 i/100ms
+Calculating -------------------------------------
+             require      6.026 (± 0.0%) i/s  (165.96 ms/i) -     31.000 in   5.145463s
+            autoload     11.730 (± 0.0%) i/s   (85.25 ms/i) -     59.000 in   5.032426s
+
+Comparison:
+             require:        6.0 i/s
+            autoload:       11.7 i/s - 1.95x  faster
+```