# Data quality for machine learning

A pack of 6 sessions from the MLOps Community YouTube channel, in the order to watch them. 3h 08m of video.
Page: https://mlopstalks.com/packs/data-quality-for-machine-learning

A table loads successfully and still gives the model the wrong evidence. A column can disappear, an event can change meaning, or a labeler can answer a different question from the one you intended. More rows will not repair those failures. Begin by defining what makes data wrong and tracing a suspicious result back through its processing. Convoy's incidents show why checking schemas alone is insufficient. GoCardless then moves expectations toward the producers, while a second contracts talk separates prevention from detection and helps decide where to invest. Labeling instructions bring the same discipline to human-created training data. Finish with a smaller, carefully checked starting dataset and a way to select useful examples, so improving quality does not require cleaning every source before you can begin.

## This pack is for you if

- Your data jobs succeed, but models receive missing, misleading or poorly understood inputs.
- Upstream application changes break training datasets without warning the people using them.
- Adding more labeled examples has not fixed inconsistent meanings or instructions.

## The talks, in order

### 1. Data Quality Management Techniques - The Complete Guide

 | 28:07 | MLOps Community
Video: https://www.youtube.com/watch?v=PfWdKVTYffo
Summary: https://mlopstalks.com/talks/data-quality-management-techniques-the-complete-guide.md

Why first: An unwelcome result is not necessarily bad data. This talk separates incorrect, incomplete and misleading records from evidence that merely challenges expectations, then works backwards through the data lifecycle. Establish that diagnostic habit before the incident accounts tempt you to prescribe a tool.

### 2. Data Quality = Quality AI

Samuel Partee, Redis & Chad Sanderson, Gable & Joe Reis, Ternary Data & Maria Zhang, Proactive AI Lab Inc & Pushkar Garg, Clari Inc | 27:15 | AIQCON 2024
Video: https://www.youtube.com/watch?v=r2fxb-OnBRM
Summary: https://mlopstalks.com/talks/data-quality-quality-ai.md

Why here: Convoy's pricing incidents expose two different breaks: a removed column and machine-generated bids entering data previously produced by people. The second can preserve the table's shape while changing its meaning. That expands the diagnosis from missing fields to source behavior, setting up the need for explicit producer commitments.

### 3. Driving ML Data Quality with Data Contracts

Andrew Jones, GoCardless | 34:30 | MLOps Meetup
Video: https://www.youtube.com/watch?v=2v2w6axyvMk
Summary: https://mlopstalks.com/talks/driving-ml-data-quality-with-data-contracts.md

Why here: Upstream teams can publish a defined data interface instead of exposing internal tables and leaving consumers to repair every change. Jones shows GoCardless's response, including ownership, versioning and migration expectations. The contract is an agreement about what downstream work may rely on, with the implementation serving that agreement.

### 4. Data Contracts: The Missing Piece of the Data Puzzle

Mark Freeman, Humu | 13:40 | MLOps Community
Video: https://www.youtube.com/watch?v=7rfWgmDRCWU
Summary: https://mlopstalks.com/talks/data-contracts-the-missing-piece-of-the-data-puzzle.md

Why here: Contracts are targeted work, so applying them everywhere is not an automatic next step. Freeman distinguishes checks on proposed changes from observation of what has already happened. Use the broader evidence to identify critical workflows, then agree on the specific changes those contracts should catch.

### 5. Data Labeling Best Practices

Charles Brecque, TextMine | 12:59 | AI in Production 2024
Video: https://www.youtube.com/watch?v=9hf_nFMiLCo
Summary: https://mlopstalks.com/talks/data-labeling-best-practices.md

Why here: People producing labels also need an unambiguous agreement. Brecque's document-extraction examples show how unclear instructions and guessed answers teach unwanted behavior. The setting is fine-tuning, but the quality question is broader: do the instructions and available evidence support the answer your training record claims?

### 6. Data Selection for Data-Centric AI: Data Quality Over Quantity

Cody Coleman | 1:11:01 | MLOps Coffee Sessions
Video: https://www.youtube.com/watch?v=v7Pj7a6KXSU
Summary: https://mlopstalks.com/talks/data-selection-for-data-centric-ai-data-quality-over-quantity.md

Why last: You do not have to repair the whole archive to begin. Coleman recommends a small clean source, checks after each transformation, and gradual expansion. His uncertain bowling images also show how example selection can reveal cases people would overlook. This turns the preceding quality requirements into a bounded first dataset and a deliberate way to improve it.

## Editor's note

The opening talk traces quality failures backwards through transformations, while the Convoy panel shows why a successful data job can still produce harmful inputs. ZenML records each pipeline run's steps, inputs, outputs and code version, giving an investigation a path back to the data and code that produced an artifact. Those records complement the source agreements and quality checks described here.

Written by the MLOps Talks editors (the ZenML team), not by any of the speakers.