MLOps Dictionary

Comprehensive terminology guide for ML solutions and feature stores.

ABCDEFGHIKLMNOPQRSTUV
Featured Terms
Core Concept

Feature Store

A data management layer that centralises the storage, transformation, and retrieval of ML features for both training and real-time inference.

Architecture

AI Lakehouse

A unified data platform that combines the scalability of a data lake with the governance of a data warehouse, purpose-built for ML workloads.

Governance

Sovereign AI

An approach where an organization builds and operates its own AI systems, ensuring data security and compliance with local regulations.

A

AI Lakehouse

AI Pipeline

Auto-regressive Models

AutoML

B

Backfill features

Backfill training data

Backpressure for feature stores

Batch Inference Pipeline

C

CI/CD for MLOps

Compound AI Systems

Context Window for LLMs

D

DAG Processing Model

Data Compatibility

Data Contract

Data Lakehouse

Data Leakage

Data Modeling

Data Partitioning

Data Pipelines

Data Quality

Data Transformation

Data Type (for features)

Data Validation (for features)

Data Warehouse

Data-Centric ML

Dimensional Modeling and Feature Stores

Downstream

E

ELT

Embedding

Encoding (for Features)

Entity

ETL

F

Feature

Feature Data

Feature Engineering

Feature Freshness

Feature Function

Feature Groups

Feature Logic

Feature Monitoring

Feature Pipeline

Feature Platform

Feature Reuse

Feature Selection

Feature Service

Feature Store: The Definitive Guide

Feature Type

Feature Value

Feature Vector

Feature View

Filtering

Fine-Tuning LLMs

Flash Attention

Function Calling with LLMs

G

Generative AI

Gradient Accumulation

Grouped Query Attention

H

Hallucinations in LLMs

Hyperparameter

Hyperparameter Tuning

I

Idempotent Machine Learning Pipelines

In Context Learning (ICL)

Inference Data

Inference Logs

Inference Pipeline

Instruction Datasets for Fine-Tuning LLMs

K

KServe

L

Lagged features

LangChain

Latent Space

LLM Code Interpreter

LLM Temperature

LLMOps

LLMs - Large Language Models

M

Machine Learning Infrastructure

Machine Learning Logs

Machine Learning Observability

Machine Learning Pipeline

Machine Learning Systems

ML

ML Artifacts (ML Assets)

MLOps

MLOps Platform

Model Architecture

Model Bias

Model Context Protocol

Model Deployment

Model Development

Model Evaluation (Model Validation)

Model Governance

Model Inference

Model Interpretability

Model Monitoring

Model Performance

Model Quantization

Model Registry

Model Serving

Model Training

Model-Centric ML

Model-Dependent Transformations

Model-Independent Transformations

Monolithic Machine Learning Pipeline

MVPS

N

Natural Language Processing (NLP)

O

Offline Store

On-Demand Features

On-Demand Transformation

Online Inference Pipeline

Online Store

Online-Offline Feature Skew

Online-Offline Feature Store Consistency

Orchestration

P

PagedAttention

Pandas UDF

Parameter-Efficient Fine-Tuning (PEFT) of LLMs

Point-in-Time Correct Joins

Precomputed Features

Prompt Engineering

Prompt Store

Prompt Tuning

Python UDF

Q

Query Engines

R

Real-Time Machine Learning

Recommender System

Representation Learning

Retrieval Augmented Generation (RAG) for LLMs

RLHF - Reinforcement Learning from Human Feedback

RoPE Scaling

S

Sample Packing

Schema

Similarity Search

Skew

Sovereign AI

Splitting Training Data

SQL UDF in Python

Streaming Feature Pipeline

Streaming Inference Pipeline

T

Test Set

Theory-of-Mind Tasks

Time travel (for features)

Train (Training) Set

Training Data

Training Pipeline

Training-Inference Skew

Transformation

Two-Tower Embedding Model

Types of Machine Learning

U

Upstream

V

Validation Set

Vector Database

Versioning (of ML Artifacts)

vLLM