PDF to MCQ generator

PDF to MCQ Converter and MCQ Generator That Turns a PDF into Multiple Choice Questions

PDFQuiz is a PDF to MCQ generator that reads a document and writes the multiple choice questions for you. Upload a PDF of a textbook chapter, lecture notes, a training manual or a study guide, and the AI pulls out the key terms, definitions and facts and turns them into MCQs with the correct answer marked and three plausible distractors. You set how many questions you want and how hard they should be, edit anything the AI drafts, then export a clean MCQ set to PDF or Word. Every question comes straight from your file, so the test covers exactly what you uploaded.

Your files are processed securely and are visible only to you. We keep them with your quizzes for as long as your account exists.

Upload your PDF and generate MCQs

Click to upload or drag and drop

PDF, DOCX, PPTX, TXT, JPG, JPEG, PNG, HEIC, ODP, ODT, BMP, or TIFF

Up to 10 files at once, 50 MB each, merged into one quiz within the free page limit

Please wait, your quiz is being created...

Uploading...

Source files
PDF, Word, slides
Question type
MCQ with answer key
Export formats
PDF and Word

The short answer on how to convert a PDF to MCQs

Upload the PDF, tell the generator how many multiple choice questions you want and how hard they should be, and it reads the document and writes each MCQ with one correct answer and three distractors drawn from your file. A typical 10 to 20 page PDF produces a full 20 question set in under a minute. You then edit any question, mark or reword the answer key, and export the MCQs to PDF or Word. Because every question comes from the text you uploaded, the test covers exactly your material instead of generic facts.

MCQ difficulty levels a PDF to MCQ generator can produce
MCQ difficulty What it tests How the stem reads Best for
Recall Does the reader remember a term, date or definition "Which term describes...", "In what year did..." Vocabulary, facts, first pass review
Comprehension Can the reader explain an idea in their own words "Which statement best summarizes...", "What does X mean here" Concepts, processes, chapter checks
Application Can the reader use a rule in a new situation "A patient presents with... which is most likely", "Given X, which applies" Exam prep, licensure, case based tests
Analysis Can the reader compare, rank or find the flaw "Which is the best explanation for...", "What is the main difference between" Higher level courses, critical reasoning

Set the difficulty before you generate. Recall MCQs are fastest to review and grade; application and analysis MCQs take longer to write by hand, which is where an automatic PDF to MCQ generator saves the most time.

What a PDF to MCQ generator does

Convert a PDF you already have into multiple choice questions

A PDF to MCQ generator takes the content inside a PDF and converts it into multiple choice questions you can answer from memory. Most documents you study or teach from already hold everything an MCQ needs: a textbook chapter, a slide deck saved as PDF, a set of class notes, a policy or a study guide. Writing four good answer options for every fact by hand is slow work. PDFQuiz reads the PDF, finds the terms, dates, names and concepts worth testing, and drafts each question with one correct answer and three believable distractors. You then take over to reword anything, swap an option, add or drop items, and set the difficulty. There is no page cap and no small file limit, so a two page handout or a two hundred page textbook both work.

Reads your PDF directly

Upload the file as is. The reader pulls the text, headings and lists out of the PDF and builds questions around the content you uploaded, not generic facts off the internet.

Real four option MCQs

Each question gets one correct answer and three distractors that are plausible enough to test real understanding, not throwaway options anyone can rule out at a glance.

Set count and difficulty

Tell the generator how many MCQs you want and whether they should test recall or application, and it matches the question style to the level you choose.

Answer key included

Every MCQ set comes with the correct answers marked, so you can self check or hand a clean copy to a class or study group without writing the key yourself.

Full editing control

Nothing is locked. Rewrite a stem, replace a weak distractor, change the wording or delete an item that misses the point. You decide the final question set before you export it.

Export to PDF or Word

Download a standalone MCQ set you own as a PDF or Word file. Print it, email it or paste it into your LMS. The questions are not trapped inside another app.

How it works

From a PDF to MCQs in three steps

1

Upload your PDF

Drop in the PDF you want to test. A Word file or slide deck works too. Combine several files if the content you want to quiz spans more than one document.

2

Generate the MCQs

Choose how many multiple choice questions you want, then let the AI read the document and draft each one with the answer key marked. It takes seconds, not an evening.

3

Edit and export

Review each question, fix anything you want to change, then export to PDF or Word to print, share or study from right away.

Every kind of PDF

What you can turn into multiple choice questions

If the content is in a PDF, it can become an MCQ set. These are the documents people convert most.

Textbook chapters

Convert a chapter PDF into MCQs that test the terms and concepts students need before an exam, then edit anything that needs tightening.

Lecture notes and slides

Turn class notes or a slide deck saved as PDF into a quick multiple choice quiz so students can check what they remember from the session.

Training manuals

Convert an onboarding or compliance PDF into a knowledge check that confirms people read and understood the material.

Study guides

Test yourself on a study guide instead of rereading it. Pair the MCQs with the study guide generator for full coverage.

Who uses it

Built for anyone who needs MCQs fast

Teachers and instructors

Convert a reading or chapter PDF into a multiple choice quiz with an answer key, then edit it to fit the class. Pairs well with the multiple choice quiz maker.

Students

Turn a textbook PDF or a stack of notes into practice MCQs before an exam, and quiz yourself instead of rereading the same pages.

Trainers and L&D

Build a quick assessment from a policy or training PDF so new hires can confirm what they learned without anyone writing four answer options by hand.

Why PDFQuiz instead of writing the MCQs by hand

The slow part of a multiple choice question is not the correct answer, it is the three wrong options that still look right. Writing those for every fact in a document takes most of an evening. PDFQuiz reads the actual content of your PDF and drafts the stem, the correct answer and the distractors together, then hands you a set you fully control. The answer key is built in, every item is editable, and you export a standalone PDF or Word file with no page cap and no small file ceiling. If you also want the same content as a broader test or as flashcards, generate them from the same upload and study more than one way.

  • MCQs generated from your own PDF, not a generic question bank.
  • Editable stems and distractors with a built in answer key on every set.
  • Export questions you own to PDF or Word, with no page or file size cap.

Three ways to turn a PDF into MCQs, compared

Writing them by hand, pasting the text into a general purpose chatbot, and using a dedicated PDF to MCQ converter all produce multiple choice questions. They differ most on two things people underestimate: whether the whole document actually gets covered, and how much editing the distractors need before the test is usable.

Comparison of methods for converting a PDF into multiple choice questions
What matters Writing MCQs by hand Pasting text into a chatbot Dedicated PDF to MCQ converter
Time for a 20 question set Two to four hours for a chapter, most of it spent on distractors Ten to twenty minutes, plus the copy and paste work per section Under a minute to draft, then ten to fifteen minutes to review
Coverage of the whole document Complete, but only because you decided the outline yourself Uneven. Long documents get summarized, and the middle is where questions go missing Spread across the file, with the question count you set applied over the document
Scanned or image only PDFs Fine, you are reading it anyway Depends on the tool. Pasting from a scan usually yields nothing Handled, since the file is read rather than pasted as text
Distractor quality Best available, because you know which wrong answers students actually pick Mixed. Implausible options and occasionally two defensible answers Drawn from the document, so the wrong options are on topic. Still needs a read through
Answer key You maintain it separately, which is where version drift starts Produced inline, usually mixed into the same block as the questions Generated as a matching key you can export separately from the student copy
Repeating it next term Reuse the same test, or start over Repeat the whole paste and prompt sequence Re-upload the revised chapter and regenerate
Where it goes wrong It does not get done, so the quiz is skipped Silent gaps in coverage that nobody notices until the exam Shipping the draft without reading the key

The review step is the part none of the three methods removes. Reading twenty generated questions against the source takes ten to fifteen minutes, and it is where you catch the item with two defensible answers before students do.

What to check before you send an MCQ set out

Four checks catch most problems. Look for questions where two options are both defensible, which is the single most common flaw in generated MCQs. Look for the giveaway pattern where the correct option is noticeably longer or more qualified than the distractors. Check that no question depends on a figure or table the reader will not have in front of them. And confirm the questions are spread across the document rather than clustered in the first few pages.

When hand writing is still the right call

For a high stakes exam that determines a grade or a certification, write the final items yourself and use a generated set as the first draft and the coverage check. Generation is strongest where volume matters and the stakes are moderate: chapter checks, practice sets, question banks, and self study. The gap between the two is almost entirely distractor quality, because knowing which wrong answer a student will reach for is knowledge about your students, not about the document.

Which PDFs produce weak MCQs, and what to fix before you upload

Open the PDF, press select all, copy, and paste into a plain text editor. What you see there is close to what any converter sees. If the pasted text reads as clean paragraphs, your questions will be good. If sentences break mid-clause or two unrelated lines are spliced together, that is the problem, and it happens before the AI ever reads a word.

The reason is structural. A PDF records where marks sit on a page, not what order a human should read them in. Reading order only exists in the file when the PDF is tagged, meaning it carries a structure tree describing headings, paragraphs, lists and tables. W3C covers this under WCAG technique PDF3. Most textbook scans, journal articles and exported handouts are untagged, and export tools that add tags automatically often get multi-column layouts wrong anyway.

What you see in the pasted text What is happening What to do about it
Sentences that switch topic halfway through A two-column page is being read straight across, so line one of the left column is joined to line one of the right column Upload the original Word or slide file instead, or re-export the PDF as a single column. This is the single biggest cause of odd questions
The same line every page or two Running heads, footers and page numbers extract as body text Harmless in small doses. On a long document it dilutes the material, so upload a page range rather than the whole book
Nothing pastes at all The page is an image, not text. Typical of scans and photographed handouts Fine to upload, since scanned pages are read directly. Keep the page flat and the whole text block in frame, because skew costs accuracy
Missing or garbled letter pairs, such as fi and fl The embedded font lacks the mapping that turns its glyphs back into characters Re-export from the source application, or print to PDF from a viewer that re-embeds the fonts
Tables arriving as a wall of numbers Cell boundaries are drawn lines, not structure, so rows and columns flatten Expect factual questions from the surrounding prose rather than from the table. State the key figures in a sentence if they must be tested
Mostly headings and bullet fragments A slide deck exported to PDF, where the meaning lived in what the speaker said Add the speaker notes, or upload the deck alongside the reading it was based on, so there are assertions to test
The upload is refused before anything is read The file is over the 50 MB per file limit. A scanned chapter photographed at 300 dpi hits that long before it runs out of pages Split the chapter you actually need out of the book, or compress the PDF at the 150 dpi level first. Avoid the strongest 72 dpi setting on a scan, because the text on those pages is an image and softening it costs OCR accuracy

How much source material a good MCQ set needs

One page of dense body prose supports about two to three solid questions. A ten page chapter comfortably carries twenty to thirty. That ratio is not a technical limit, it is where the material runs out: past it the generator has no fresh assertions left and starts testing trivia or rewording facts it already used, which is how near-duplicate items appear in a set.

If you need more questions than the ratio allows, add source rather than raising the count. Upload the chapter together with the lecture notes covering it, and the extra angles give the generator somewhere new to go. The reverse also holds: on a very long upload, narrowing to the pages you are actually being tested on produces sharper questions than the whole book does, because everything the model reads competes for the same twenty slots.

Reference on PDF reading order and tagging: W3C, WCAG technique PDF3, Ensuring correct tab and reading order in PDF documents.

How many answer options should each MCQ have?

Three. The research is clearer than most people expect. Rodriguez (2005), a meta-analysis in Educational Measurement: Issues and Practice pooling roughly eighty years of studies, found that three-option multiple choice items perform as well psychometrically as four- and five-option items, while taking less time to answer. Fewer, better options let you fit more questions into the same test, which improves how much of the material you actually cover.

The reason is that most fourth and fifth options are dead weight. A distractor only works if a real student, holding a real misconception, would actually pick it. In practice most items have two or three genuinely tempting wrong answers and then a throwaway that nobody chooses. That throwaway adds length without adding measurement.

Options per item Random guess rate When to use it
Three33%The default. Matches four and five options on quality and lets you ask more questions in the same time.
Four25%Fine when you genuinely have three plausible misconceptions, common on high stakes or certification style items.
Five20%Rarely worth it. The fifth option is usually filler that lengthens the test without measuring anything.

There is a one-step edit that puts this to work on any generated draft: read each item and delete the option you would not have written yourself. If you would keep all four, you had three real misconceptions and the fourth earned its place. If one option feels like filler, it was, and cutting it makes the item both shorter and sharper.

What question types can it generate besides MCQs?

Six, and you pick which ones are switched on before the file is processed. Multiple choice is only the default. The other five matter when a single MCQ set is not enough to test the material: recall questions need a blank rather than options, and paired terminology is tested far better by matching than by four lookalike choices.

The shapes below are the exact specification the generator works to, not a marketing summary. Each type returns a fixed structure, which is why the output imports cleanly into other systems.

Question type What comes back Best for
Single choiceExactly 4 options, a to d, with exactly 1 correctThe standard MCQ, and the default this page is about
Multiple answerExactly 4 options, with 2 or 3 of them correctMaterial where more than one thing is true and partial knowledge should not pass
True or falseExactly 2 optionsFast coverage of a long document, and checking for common misconceptions
Fill in the blankA sentence containing a blank, plus the exact text that fills it. No options at allTerminology, dates and figures, where recognising an option is too easy
MatchingA set of paired items, minimum 4 pairs per questionVocabulary, definitions, drug names, statute numbers, anything learned in pairs
Open endedA written question plus a model answer, phrased so it never hints at optionsChecking whether someone can explain the idea rather than recognise it

One behaviour is worth knowing before you set the count. The total is split evenly across whichever types you enabled, so asking for 30 questions with three types switched on returns 10 of each rather than 30 MCQs with a few extras. If you want a pure MCQ set, leave single choice as the only type enabled.

How do you know if the generated MCQs are any good?

Cover the answer options and try to answer the question from the stem alone. If you can, the item is testing whether someone learned the material. If you cannot answer until you read the options, the stem is too vague and the item is really testing reading speed. That single check is the cover-the-options rule from the National Board of Medical Examiners item-writing guide, and it catches most of the weak questions in any generated set in about ten seconds each.

The same guide is useful for a second reason. It lists the specific technical flaws that let a student who does not know the answer pick the right option anyway. These are worth knowing because they are the flaws an automated MCQ maker is most likely to produce, and because every one of them is a fast edit rather than a rewrite. Run a generated draft against this list before you give the test to anyone.

Technical item flaws that give away the correct answer, and how to fix each one
Flaw in the item Why it hands over the answer The edit
The correct option is the longest Writers explain the right answer and leave the wrong ones short. Students learn to pick the long one without reading it. Cut the explanation out of the key so all options run roughly the same length.
Absolute terms in the distractors Options containing "always" or "never" read as false on sight, so they get eliminated without any knowledge of the topic. Delete the absolute term or drop the option and let the item run with three.
Grammatical mismatch An option that does not follow grammatically from the stem, singular where the stem needs a plural, is visibly not the intended answer. Make every option singular or every option plural, and close the stem with a full question.
Word repeat, the clang clue A distinctive word from the stem reappears only in the correct option, which points straight at it. Reword the stem, or use the repeated word in all of the options.
"None of the above" It tests whether the student can rule out three statements, not whether they know the actual answer. Replace it with a specific option, such as a concrete action or a real value.
A stem phrased as "all of the following EXCEPT" On a test where every other item is positive, students miss the word "except" even in bold, and lose the point for the wrong reason. Rewrite the stem positively and ask directly for the one thing you want.
Vague frequency words "Often" and "usually" mean different things to different readers, so the item measures interpretation instead of knowledge. Put a number in, or cut the qualifier entirely.
Options that cover every possibility When a subset of the options is collectively exhaustive, the answer can be reasoned out by elimination alone. Swap at least one option out, and take care not to leave an obvious pair behind.

Flaw categories and the recommended fixes follow the National Board of Medical Examiners, Item-Writing Guide, chapters 3 and 4.

A distractor does not have to be completely wrong

This is the piece most people writing their first MCQ set get backwards, and it is worth knowing before you edit a generated draft. In a one-best-answer question the task is to pick the most correct option, not the only correct one. The NBME guide is explicit about it: as long as the options sit on a single continuum, from least likely to most likely, the distractors do not have to be totally wrong. A question where three options are obviously absurd tests nothing, because a student who read the chapter once will still get it right.

So when you review the draft, resist the urge to make wrong answers more wrong. The better move is the opposite: make them more tempting. Ask what someone who half learned this material would believe, and put that in an option. A distractor earns its place when a real misconception would lead a real student to choose it.

Which questions to cut after you have used the test once

Two numbers tell you almost everything about an item once you have shared the quiz with a class or run it with a team. PDFQuiz does not calculate either one, so this is a description of what to work out for yourself from the answer sheets, not a feature list. The first is the percentage of people who got the item right, what test developers call the p-value. Items above 95 percent or below 30 percent tell you very little about the group, because nearly everyone or nearly nobody got them. They are not automatically bad questions, but a test built mostly out of them cannot separate people who learned the material from people who did not.

The second is the item-total correlation, the relationship between getting one question right and scoring well on the test overall. It runs from minus 1.0 to plus 1.0, and you want it positive: the people who did well on the whole test should be the people who got this item right. A negative correlation is the useful signal, because it usually means the answer key is wrong. The strongest students picked an option you marked incorrect, which is exactly what happens when a generated item was keyed to the wrong choice or the source passage was ambiguous. Check those items first.

If you use PDFQuiz as an MCQ creator for the same course every term, this is where a generated bank starts to beat a handwritten one. Regenerating from the same PDF gives you fresh items to swap in for the ones that did not discriminate, and you keep the questions that did.

Can you convert a PDF that already contains MCQs into a quiz?

Yes, and it is one of the most common uses of this converter, but be clear about what happens. PDFQuiz is a generator rather than an importer. It reads the question bank you upload and writes multiple choice questions covering the same material, in interactive form with the answer key marked. Expect close coverage of your bank rather than a guaranteed line by line copy of it, and check the set against the original before you assign it.

In practice that distinction matters most for two things. First, numbering: if your PDF question bank is ordered so that question 14 refers back to question 13, upload those questions together and read the generated pair, because a generator has no obligation to keep them adjacent. Second, wording you are contractually stuck with. If you are working from a licensed item bank whose phrasing cannot change, generation is the wrong tool and you want to retype those items into the editor by hand, where nothing rewrites them.

What you get back either way is the part a PDF cannot do: a quiz that grades itself. A question bank sitting in a PDF has to be printed, sat, and marked by a person. The same bank uploaded here becomes a link you send out, taken on a phone or a laptop, with each attempt and score landing in your dashboard. For a tutor with ten years of past papers on disk, that is the whole reason to convert them.

Converting old exam papers, in order

Work one paper at a time rather than merging five years into a single upload. Each generation accepts up to 10 files of 50 MB each, so a set of past papers fits comfortably, but keeping them separate means you can name each quiz after its year and see which topics students still miss. Merged, you lose that entirely and you get one long undifferentiated set.

Scanned papers work, they just need the OCR option turned on, and the quality of the scan decides the quality of the questions. A crisp 300 dpi scan of a printed paper reads cleanly. A phone photo taken at an angle under a desk lamp does not, and the questions will show it. If your scans are large enough that the uploader refuses them, compress the PDF before you retry rather than rescanning at a lower resolution, because on a scan the text is part of the image and dropping resolution is what costs you accuracy.

What changes when you turn a PDF into a mock test, question paper or exam

The generation step is identical for all three. What changes is the settings you apply afterwards, and it takes about a minute. The same PDF becomes a mock test by adding a timer and shuffling the order, a question paper by exporting to PDF or Word with the answer key on a separate page, or a graded exam by sharing one link and reading the scores as they come in.

How the same generated MCQ set is configured as a mock test, a printed question paper or a graded exam
You want Question mix that works What you turn on How it ends up
A mock test Single choice MCQs almost throughout, so the timing matches the real paper The timer, and shuffled question order so a second attempt is not the same run A link the candidate sits under exam conditions, scored the moment they finish
A printed question paper MCQs plus a few open ended questions, since a person is marking anyway Export to PDF or Word, then keep the answer key as a separate handout A paper you photocopy, with no devices needed in the room
A graded exam Mostly MCQs, with fill in the blank for anything that must be recalled rather than recognized Share to your class or team so each attempt is tied to a name Scores per person in the dashboard, with the questions most people missed visible at a glance

One thing worth deciding up front: a mock test should be harder than the real one or it teaches false confidence. The cheapest way to do that here is not to raise the difficulty setting but to keep the timer tight and to leave in the questions candidates got wrong last time. If you want the full walkthrough for exam rehearsal specifically, the PDF to practice test page goes through it, and PDF to test covers the graded classroom version.

Frequently asked questions about converting a PDF to MCQs

How do I convert a PDF to MCQs?
Upload your PDF on this page and click generate. The AI reads the document, pulls out the key terms and facts, and writes multiple choice questions with one correct answer and three distractors, plus an answer key, in seconds. You can then edit any question and export the set to PDF or Word. There is nothing to install and no formatting work on your end.
Can AI generate MCQs from a PDF?
Yes. PDFQuiz uses AI to read the content of your PDF and generate multiple choice questions based on what the document actually says, including definitions, dates, names and concepts. It writes the stem, marks the correct answer and drafts three plausible distractors, then leaves every item open for you to edit before you export it.
How many MCQs can I generate from one PDF?
You set the number. A rough guide is three to five solid MCQs per page of substantive content, so a ten page PDF might yield thirty to fifty questions and a full textbook chapter many more. There is no page cap, so you can convert a long document at once and regenerate for a fresh batch later to grow your question bank.
Can I edit the multiple choice questions it generates?
Yes, every part of every question is editable. Rewrite the stem, replace a distractor that is too easy to rule out, change the correct answer, adjust the wording or delete an item entirely. The AI gives you a strong first draft and you keep full control over the final MCQ set before exporting.
Does the PDF to MCQ converter work with scanned PDFs?
It works best with PDFs that contain real, selectable text. A scanned PDF is a picture of a page rather than selectable text, so it has to be read visually, and the generator does that directly. You do not need an external OCR tool or to retype anything first: turn the OCR option on and upload the scan as it is. Legibility is what decides the result: a clean scan or a straight, well lit photo produces solid MCQs, while a dark or skewed one produces fewer.
Can I export the MCQs to Word, Canvas or Moodle?
PDFQuiz exports your multiple choice questions and answer key to PDF or Word, not to a Canvas or Moodle import package. To use them in your LMS, generate the set here, then paste the questions into your LMS quiz builder or attach the exported file. The hard part, writing the questions from your PDF, is already finished.
Why did my PDF produce bad multiple choice questions?
Usually the text came out of the PDF in the wrong order. A two-column academic paper extracts as interleaved lines unless the file is tagged, so sentences from the left and right columns get spliced together and the generator writes questions from text that was never a sentence. Copy the PDF into a plain text editor to see exactly what a converter sees, then export a single-column copy or upload the source Word file instead.
How many pages of PDF do you need for a good MCQ set?
About one page of dense body prose per two to three questions. A ten page chapter comfortably supports twenty to thirty items. Asking for forty questions from three pages forces the generator to test trivia and near-duplicates, because it runs out of assertions worth testing and starts rewording the same fact in a new stem.
What makes a good multiple choice question?
A good MCQ has a clear stem, one defensibly correct answer, and distractors that are plausible enough that someone who has not learned the material could believe them. It should test understanding rather than trick the reader. PDFQuiz drafts each question to that standard, and you can sharpen any distractor that reads as obviously wrong.
Which parts of a PDF make the best MCQs?
Body prose that states definitions, causes, procedures, and comparisons produces the strongest items. Tables and figure captions produce good factual questions. What produces weak questions is front matter, reference lists, and pages that are mostly headings, because there is no assertion to test. If a chapter opens with ten pages of introduction, upload only the section you actually taught.
Can you make a printed test from the generated MCQs?
Yes. Export the set to PDF and print it, keeping the answer key as its own file rather than a hidden column in the same document. Roughly ten to twelve four option items fit on a US letter page at 11 point with normal margins. The printable test maker covers the page layout, and the answer key generator handles the separate key.

Convert your PDF to MCQs now

Upload your PDF, generate multiple choice questions with an answer key, edit them, and export to PDF or Word.

Generate your MCQs

Related tools and guides

PDF to quiz

Convert any PDF into a full quiz with mixed question types and an answer key.

Multiple choice quiz maker

Build a multiple choice quiz from any source with smart distractors.

MCQ generator

Generate multiple choice questions from text, notes or a document.

PDF question generator

Generate questions of every type from a PDF, not just MCQs.

AI quiz generator

Generate a quiz from any document with AI.

Notes to quiz

Convert class or lecture notes into a practice quiz.

Popular certification and exam practice tests

Upload your own study notes, review book, or exam handbook and PDFQuiz builds practice questions for the test you are preparing for. Jump straight to the practice test for your exam:

Cloud (AWS)

Cloud (Google Cloud)

DevOps and infrastructure

Data platform

Cloud (Microsoft Azure)

Finance and accounting

College and grad admissions