Add support for OpenAI function calling inputs in chat UI parsing by daniellok-db · Pull Request #20058 · mlflow/mlflow

daniellok-db · 2026-01-16T08:15:36Z

Related Issues/PRs

#20054

What changes are proposed in this pull request?

The chat UI was not properly recognizing OpenAI's function calling inputs as chat messages. For example, when the input array in a span contains items like:

{
  "input": [
    {"role": "user", "content": "What is my horoscope?"},
    {"type": "reasoning", "id": "...", "summary": []},
    {"type": "function_call", "call_id": "...", "name": "get_horoscope", "arguments": "..."},
    {"type": "function_call_output", "call_id": "...", "output": "..."}
  ]
}

The UI would fail to parse these as chat messages because normalizeOpenAIResponsesInput only recognized OpenAIResponsesInputMessage items (with role and content).

This PR extends it to accept other formats such as function_call and function_call_output by reusing the existing isOpenAIResponsesOutputItem helper

How is this PR tested?

New unit/integration tests

Added new test cases:

handles input with function calls and function call outputs

The trace was generated by following the official OpenAI function calling tutorial:

https://platform.openai.com/docs/guides/function-calling

Before, no chat messages appear:


Summary view	Detail view

After:


Summary view	Detail view

Does this PR require documentation update?

No. You can skip the rest of this section.

Release Notes

Is this a user-facing change?

Yes. Give a description of this change to be included in the release notes for MLflow users.

Fixed an issue where the tracing UI would not properly display OpenAI function calling inputs as chat messages.

What component(s), interfaces, languages, and integrations does this PR affect?

Components

area/uiux: Front-end, user experience, plotting, JavaScript, JavaScript dev server
area/tracing: MLflow Tracing features, tracing APIs, and LLM tracing functionality

How should the PR be classified in the release notes? Choose one:

rn/bug-fix - A user-facing bug fix worth mentioning in the release notes

Should this PR be included in the next patch release?

Yes (this PR will be cherry-picked and included in the next patch release)
No (this PR will be included in the next minor release)

🤖 Generated with Claude Code

github-actions · 2026-01-16T08:15:58Z

🛠 DevTools 🛠

Install mlflow from this PR

# mlflow
pip install git+https://github.com/mlflow/mlflow.git@refs/pull/20058/merge
# mlflow-skinny
pip install git+https://github.com/mlflow/mlflow.git@refs/pull/20058/merge#subdirectory=libs/skinny

For Databricks, use the following command:

%sh curl -LsSf https://raw.githubusercontent.com/mlflow/mlflow/HEAD/dev/install-skinny.sh | sh -s pull/20058/merge

Copilot

Pull request overview

This PR extends the chat UI parsing to properly recognize OpenAI function calling inputs that appear in the input array alongside regular chat messages. Previously, the UI only recognized OpenAIResponsesInputMessage items (with role and content), but OpenAI's Responses API can include additional item types like function_call, function_call_output, and reasoning in the input array.

Changes:

Added support for parsing function call and function call output items in OpenAI Responses input arrays
Extended reasoning attachment logic to work with input arrays
Added comprehensive unit tests for the new functionality

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
mlflow/server/js/src/shared/web-shared/model-trace-explorer/chat-utils/openai.ts	Added `isOpenAIResponsesInputItem`, `normalizeOpenAIResponsesInputArrayItem`, and `processInputItemsWithReasoning` functions to handle mixed input arrays containing messages, function calls, and reasoning items
mlflow/server/js/src/shared/web-shared/model-trace-explorer/chat-utils/openai.test.ts	Added two test cases covering function calls with/without reasoning summaries in input arrays

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

mlflow/server/js/src/shared/web-shared/model-trace-explorer/chat-utils/openai.ts

mlflow/server/js/src/shared/web-shared/model-trace-explorer/chat-utils/openai.test.ts

github-actions · 2026-01-16T08:23:30Z

Documentation preview for d0f9231 is available at:

https://pr-20058--mlflow-docs-preview.netlify.app/docs/latest/

More info

Ignore this comment if this PR does not change the documentation.
The preview is updated when a new commit is pushed to this PR.
This comment was created by this workflow run.
The documentation was built by this workflow run.

Signed-off-by: Daniel Lok <daniel.lok@databricks.com>

daniellok-db · 2026-01-22T13:25:06Z

mlflow/server/js/src/shared/web-shared/model-trace-explorer/chat-utils/openai.ts

+  if (
+    isArray(input) &&
+    // openai inputs can consititute of output items such as function calls and function call outputs
+    input.every((message: unknown) => isOpenAIResponsesInputMessage(message) || isOpenAIResponsesOutputItem(message))


i was going to implement some separate functions for detecting if we have function_call or function_call_output but realized it already exists in isOpenAIResponsesOutputItem.

it also does not seem to pose a problem to use normalizeOpenAIResponsesOutputItem to normalize the input as we do want them to be formatted as assistant messages (since the assistant is the one that requests the tools)

serena-ruan

LGTM!

…flow#20058) Signed-off-by: Daniel Lok <daniel.lok@databricks.com>

…0058) Signed-off-by: Daniel Lok <daniel.lok@databricks.com>

Copilot AI review requested due to automatic review settings January 16, 2026 08:15

github-actions bot added the v3.9.0 label Jan 16, 2026

github-actions bot added area/tracing MLflow Tracing and its integrations area/uiux Front-end, user experience, plotting, JavaScript, JavaScript dev server rn/bug-fix Mention under Bug Fixes in Changelogs. labels Jan 16, 2026

Copilot started reviewing on behalf of daniellok-db January 16, 2026 08:16 View session

Copilot AI reviewed Jan 16, 2026

View reviewed changes

fix

d0f9231

Signed-off-by: Daniel Lok <daniel.lok@databricks.com>

daniellok-db force-pushed the chat-ui-fix branch from db6bf4e to d0f9231 Compare January 22, 2026 13:22

daniellok-db commented Jan 22, 2026

View reviewed changes

daniellok-db added the team-review Trigger a team review request label Jan 22, 2026

github-actions bot requested review from B-Step62 and serena-ruan January 22, 2026 13:28

serena-ruan approved these changes Jan 23, 2026

View reviewed changes

github-actions bot assigned serena-ruan Jan 23, 2026

daniellok-db added this pull request to the merge queue Jan 25, 2026

Merged via the queue into mlflow:master with commit d21f573 Jan 25, 2026
60 of 62 checks passed

daniellok-db deleted the chat-ui-fix branch January 25, 2026 07:06

harupy pushed a commit to harupy/mlflow that referenced this pull request Jan 28, 2026

Add support for OpenAI function calling inputs in chat UI parsing (ml…

1e7d51d

…flow#20058) Signed-off-by: Daniel Lok <daniel.lok@databricks.com>

harupy pushed a commit to harupy/mlflow that referenced this pull request Jan 28, 2026

Add support for OpenAI function calling inputs in chat UI parsing (ml…

eb039d9

…flow#20058) Signed-off-by: Daniel Lok <daniel.lok@databricks.com>

harupy pushed a commit that referenced this pull request Jan 28, 2026

Add support for OpenAI function calling inputs in chat UI parsing (#2…

b323334

…0058) Signed-off-by: Daniel Lok <daniel.lok@databricks.com>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add support for OpenAI function calling inputs in chat UI parsing#20058

Add support for OpenAI function calling inputs in chat UI parsing#20058
daniellok-db merged 1 commit intomlflow:masterfrom
daniellok-db:chat-ui-fix

daniellok-db commented Jan 16, 2026 •

edited

Loading

Uh oh!

github-actions bot commented Jan 16, 2026

Install mlflow from this PR

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions bot commented Jan 16, 2026 •

edited

Loading

Uh oh!

daniellok-db Jan 22, 2026

Uh oh!

serena-ruan left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

daniellok-db commented Jan 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Related Issues/PRs

What changes are proposed in this pull request?

How is this PR tested?

Does this PR require documentation update?

Release Notes

Is this a user-facing change?

What component(s), interfaces, languages, and integrations does this PR affect?

How should the PR be classified in the release notes? Choose one:

Should this PR be included in the next patch release?

Uh oh!

github-actions bot commented Jan 16, 2026

Install mlflow from this PR

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions bot commented Jan 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

daniellok-db Jan 22, 2026

Choose a reason for hiding this comment

Uh oh!

serena-ruan left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

daniellok-db commented Jan 16, 2026 •

edited

Loading

github-actions bot commented Jan 16, 2026 •

edited

Loading