[DISCUSSION] DataFusion Road Map: Q1 2026

Introduction

DataFusion is a community project, and unlike many other open source projects there is no "DataFusion Company" that pays contributors and determines their priority. The project's evolution and priorities is determined by the community members able and willing to devote their time and energy to driving the project forward (PRs, code review, planning, etc)

As we have done in the past, I think it is helpful to have a public discussion about what features people are planning to work on to:

Find others interested in the same features (to combine efforts and collaborate)
Find areas of broad community interest (areas that many users want are more likely to get the requisite focus to get done)

Prior Art

Per the Roadmap Section in the docs:

Epics offer a high level roadmap of what the DataFusion community is thinking about.

Community Request

Please create or leave comments on proposal epics you are interested in helping with (to help drive / build momentum). Commenting on this ticket is also great if you are looking to help out.
Comment below about what you are planning to help with / work on

Context

Previous discussions:

Update roadmap links for DataFusion Q1 2026

I personally plan to focus in Q1 on performance

[EPIC] Parquet filter pushdown into scan #3462
[EPIC] ListingTable object store usage improvements #17214
[EPIC] Make DataFusion the top of the ClickBench Parquet leaderboard #18489
[EPIC] Support VARIANT type for unstructured data #16116 (and related technical areas like expr pushdown)
Make adding an extension type a pleasant experience (epic TBF, see here)

There was also a demand of supporting lambda syntax like https://duckdb.org/docs/stable/sql/functions/list#list_filterlist-lambdax to unblock mostly array related builtin functions, not sure if it fits in Q1, but I probably try to invest some time in there.

There was also a demand of supporting lambda syntax like https://duckdb.org/docs/stable/sql/functions/list#list_filterlist-lambdax to unblock mostly array related builtin functions, not sure if it fits in Q1, but I probably try to invest some time in there.

A somewhat similar lambda PR timed out recently - #17220

What I'd like to see worked on and what I myself am capable of working on are unfortunately not aligned :/ That being said I myself am likely to

take another stab at logical planning performance issues Planning time for queries with many columns with union and order by is very slow #17261
perhaps look into what it would take to implement something similar to Spark's flatMapGroups
additional improvements for timestamp with tz functionality datafusion.execution.time_zone is not used for basic time zone inference #13212 (comment)

Update roadmap links for DataFusion Q1 2026 (#18495)

Update roadmap links for DataFusion Q1 2026 (apache#18495)

I would love to get #11745 (support for pushing down parquet struct field access from projections and filters into scans) across the line in Q1. I've been working on the issue for months and it is our system's top priority.

all i want is to get the filenames when i read a folder of csv, json etc :)

all i want is to get the filenames when i read a folder of csv, json etc :)

I think there's pretty good precedent for that - maybe something like apache/arrow-rs#9117 using an extension type marker but in DataFusion e.g. in ParquetOpener or `FileScanConfig? I recommend you find of open issue or open one and maybe edit your comment above to attach it.

all i want is to get the filenames when i read a folder of csv, json etc :)

There is an open issue here: #6051

I would love running the same challenge for DataFusion https://www.confessionsofadataguy.com/duckdb-beats-polars-for-1tb-of-data/

#19815

Hi, i've been lurking for good new comers issues, but i couldn't figure out a good starting point to help !

thanks in advance for the help

@AndreaBozzo

Thanks @AndreaBozzo -- sadly they are hard to come by these days. I would recommend looking at some of the open PRs and helping try and move them along

Async UDTFs for remote table providers? (#10889, #16520)

Hi all 👋

I’m interested in contributing in Q1, especially around performance and Parquet pushdown improvements.

I’ll start by reviewing related issues/PRs (e.g., #17261 and Parquet EPICs).
Happy to help where needed — any guidance is appreciated

@SKM2227229725

Hi @SKM2227229725 -- perhaps you have some time to do analysis on some of the queries here:

[EPIC] Fix performance regressions when enabling parquet filter pushdown (late materialization) #20324

We also need to think on #20919 looks like depending on the batch_size the behavior can be unstable.

@neilconway

I have been thinking with @neilconway @gabotechs and others that we might want to start spending more time on coordination / consensus building. Things are the moment are pretty ad hoc / hard to understand what is going on / what the priorites are. Now that we have so many downstream stakeholders it might be time to improve the situatio

Hey all,

Q1 is almost over, but I’d still love to get involved and help where I can over the next couple of weeks.

I’m not trying to force a big roadmap item in a short window. I’d be happy to contribute wherever there’s a good fit, especially on something scoped enough to make real progress quickly, whether that’s fixing a smaller issue, improving contributor/development workflow, or helping move a larger effort forward through repros, benchmarks, or root-cause analysis.

For context, I recently spent some time reproducing and writing up the CI behavior in #19095, and I enjoyed that kind of work too.

Longer term I’m interested in getting deeper into the codebase, especially around planner/performance topics, but for the rest of Q1 I’m mostly looking to be helpful, build context, and contribute where an extra set of hands would be useful.

@adp2201

@adp2201 has filed a new discussion:

Q2 2026 roadmap planning #21017

@alamb

Hey all ,
I think it would be great doing to get the docs head of the line :) . feel free to tag anywhere .
CC: @alamb , @xudong963 @adriangb

Let's close this one down and use the new discussion:

Q2 2026 roadmap planning #21017

Navigation Menu

Repository navigation

[DISCUSSION] DataFusion Road Map: Q1 2026 #18494

Introduction

Prior Art

Community Request

Context

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

[DISCUSSION] DataFusion Road Map: Q1 2026 #18494

Description

Introduction

Prior Art

Community Request

Context

Activity

alamb commented on Nov 5, 2025

comphead commented on Nov 5, 2025

Omega359 commented on Nov 5, 2025

Omega359 commented on Nov 5, 2025

adriangb commented on Jan 13, 2026

djouallah commented on Jan 14, 2026

adriangb commented on Jan 14, 2026

Jefffrey commented on Jan 14, 2026

comphead commented on Jan 14, 2026

AndreaBozzo commented on Jan 19, 2026

alamb commented on Jan 19, 2026

aditanase commented on Jan 22, 2026

SKM2227229725 commented on Feb 11, 2026

alamb commented on Feb 13, 2026

comphead commented on Mar 12, 2026

alamb commented on Mar 13, 2026

adp2201 commented on Mar 16, 2026

alamb commented on Mar 18, 2026

Adez017 commented on Apr 24, 2026

alamb commented on Apr 27, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions