Historical DP-100 Scope
DP-100 was Microsoft's Designing and Implementing a Data Science Solution on Azure exam until its retirement on 2026-06-01. Its final published scope covered Design and prepare a machine learning solution, Explore data, and run experiments, Train and deploy models, Optimize language models for AI applications. This page preserves that dated outline for historical comparison and curriculum cleanup. It is not a schedulable exam, and its prices, delivery details, registration links, and practice calls to action must not be used for a new certification plan.
Who DP-100 Was For
Nobody should start a new certification plan around DP-100. This page serves former candidates, credential holders, training teams removing stale material, and practitioners deciding whether prior preparation is reusable for AI-300. New candidates should verify the active Microsoft page and current dated study guide before booking.
Skills You Should Be Ready to Demonstrate
- Preserve practical experience from Design and prepare a machine learning solution only where the current guide shows meaningful overlap
- Preserve practical experience from Explore data, and run experiments only where the current guide shows meaningful overlap
- Preserve practical experience from Train and deploy models only where the current guide shows meaningful overlap
- Preserve practical experience from Optimize language models for AI applications only where the current guide shows meaningful overlap
How to Reuse Your Preparation
Do not rename an old DP-100 checklist or question bank as AI-300. Compare the two official guides objective by objective and classify each historical skill as transferable, changed, or absent. Use transferable experience as a diagnostic baseline, then rebuild labs and review around AI-300's current domain weights, products, role expectations, security controls, and operational tasks. Because Microsoft identifies AI-300 as the direct replacement, review the skill map below before assuming equivalence.
What changed from DP-100 to AI-300
| Historical DP-100 | Current AI-300 | Practical impact |
|---|
Design and prepare a machine learning solution
20-25% | Design and implement an MLOps infrastructure
15-20% | The successor shifts from project setup toward reproducible infrastructure, source control, pipelines, environments, security, and governance. |
Explore data and run experiments; train and deploy models
45-55% combined | Implement machine learning model lifecycle and operations
25-30% | Experimentation remains useful, but AI-300 emphasizes registration, deployment, monitoring, retraining, lineage, and operational response. |
Optimize language models for AI applications
25-30% | GenAIOps infrastructure, quality assurance, observability, and optimization
40-55% combined | The replacement expands model optimization into production evaluation, tracing, safety, performance, RAG quality, and agent operations. |
Migration checklist
- Stop scheduling, selling, or presenting DP-100 as a current exam.
- Compare completed DP-100 preparation with the current AI-300 guide; do not carry over obsolete weighting or mechanics.
- Create a fresh AI-300 plan around its current domains and hands-on role expectations.
- Remove or quarantine practice content that cannot be mapped to a current objective with source evidence.
- Recheck Microsoft's dated AI-300 study guide before booking because role-based exams change over time.
Open the official AI-300 study guide
Historical Domain Guide
Design and prepare a machine learning solution: Historical Scope
Historically, Design and prepare a machine learning solution required candidates to connect configuration choices with operational outcomes. The official outline included Design a machine learning solution; Create and manage a workspace. Preserve hands-on understanding of those tasks where the successor comparison shows overlap, but do not treat this retired weighting or terminology as a current exam blueprint.
- Use the official historical objectives to document what design and prepare a machine learning solution required
- Mark each prior skill as transferable, changed, or absent in the current replacement guide
- Rebuild practical exercises around current services, interfaces, governance, and role expectations
Explore data, and run experiments: Historical Scope
Historically, Explore data, and run experiments required candidates to connect configuration choices with operational outcomes. The official outline included Use automated machine learning to explore optimal models; Select and understand training options. Preserve hands-on understanding of those tasks where the successor comparison shows overlap, but do not treat this retired weighting or terminology as a current exam blueprint.
- Use the official historical objectives to document what explore data, and run experiments required
- Mark each prior skill as transferable, changed, or absent in the current replacement guide
- Rebuild practical exercises around current services, interfaces, governance, and role expectations
Train and deploy models: Historical Scope
Historically, Train and deploy models required candidates to connect configuration choices with operational outcomes. The official outline included Run model training scripts; Use logs to troubleshoot job run errors. Preserve hands-on understanding of those tasks where the successor comparison shows overlap, but do not treat this retired weighting or terminology as a current exam blueprint.
- Use the official historical objectives to document what train and deploy models required
- Mark each prior skill as transferable, changed, or absent in the current replacement guide
- Rebuild practical exercises around current services, interfaces, governance, and role expectations
Optimize language models for AI applications: Historical Scope
Historically, Optimize language models for AI applications required candidates to connect configuration choices with operational outcomes. The official outline included Prepare for model optimization; Optimize through prompt engineering and prompt flow. Preserve hands-on understanding of those tasks where the successor comparison shows overlap, but do not treat this retired weighting or terminology as a current exam blueprint.
- Use the official historical objectives to document what optimize language models for ai applications required
- Mark each prior skill as transferable, changed, or absent in the current replacement guide
- Rebuild practical exercises around current services, interfaces, governance, and role expectations
Sources and Verification
Verified 2026-08-26
How this page was made
Cert Atlas compared Microsoft's final DP-100 study guide with the current AI-300 guide at the domain and objective level, then checked Microsoft's retirement list and transition announcement. OpenAI Codex assisted with extraction, normalization, comparison, and drafting. Claims were reviewed against the linked official sources; no exam stems, answers, choices, or explanations were used.
Historical DP-100 Domains
1.0 Design and prepare a machine learning solution
20-25%
- 1.1Design a machine learning solution
Identify the structure and format for datasets; Determine the compute specifications for machine learning workload; Select the development approach to train a model; Create and manage resources in an Azure Machine Learning workspace
- 1.2Create and manage a workspace
Create and manage datastores; Create and manage compute targets; Set up Git integration for source control; Create and manage assets in an Azure Machine Learning workspace
- 1.3Create and manage data assets
Create and manage environments; Share assets across workspaces by using registries
2.0 Explore data, and run experiments
20-25%
- 2.1Use automated machine learning to explore optimal models
Use automated machine learning for tabular data; Use automated machine learning for computer vision; Use automated machine learning for natural language processing
- 2.2Select and understand training options
Select and understand training options, including preprocessing and algorithms; Evaluate an automated machine learning run, including responsible AI guidelines
- 2.3Use notebooks for custom model training
Use the terminal to configure a compute instance; Access and wrangle data in notebooks; Wrangle data interactively with attached Synapse Spark pools and serverless Spark compute; Retrieve features from a feature store to train a model
- 2.4Track model training by using MLflow
Evaluate a model, including responsible AI guidelines; Automate hyperparameter tuning; Select a sampling method; Define the search space; Define the primary metric; Define early termination options
3.0 Train and deploy models
25-30%
- 3.1Run model training scripts
Consume data in a job; Configure compute for a job run; Configure an environment for a job run; Track model training with MLflow in a job run; Define parameters for a job; Run a script as a job
- 3.2Use logs to troubleshoot job run errors
- 3.3Implement training pipelines
Create custom components; Create a pipeline; Pass data between steps in a pipeline; Run and schedule a pipeline; Monitor and troubleshoot pipeline runs
- 3.4Manage models
Define the signature in the MLmodel file; Package a feature retrieval specification with the model artifact; Register an MLflow model; Assess a model by using responsible AI principles
- 3.5Deploy a model
Configure settings for online deployment; Deploy a model to an online endpoint; Test an online deployed service; Configure compute for a batch deployment; Deploy a model to a batch endpoint; Invoke the batch endpoint to start a batch scoring job
4.0 Optimize language models for AI applications
25-30%
- 4.1Prepare for model optimization
Select and deploy a language model from the model catalog; Compare language models using benchmarks; Test a deployed language model in the playground
- 4.2Optimize through prompt engineering and prompt flow
Test prompts with manual evaluation; Define and track prompt variants; Create prompt templates; Define chaining logic with the prompt flow SDK; Use tracing to evaluate your flow
- 4.3Optimize through Retrieval Augmented Generation (RAG)
Prepare data for RAG, including cleaning, chunking, and embedding; Configure a vector store; Configure an Azure AI Search-based index store; Evaluate your RAG solution
- 4.4Optimize through fine-tuning
Prepare data for fine-tuning; Select an appropriate base model; Run a fine-tuning job; Evaluate your fine-tuned model