How I Used AI Flashcards to Pass the SIE, Series 63, and Series 7

By Steven Urry

Over the past several weeks I prepared for and passed the SIE, the Series 63, and the Series 7. My study process combined two tools: Achievable for the course content and practice exams, and Memgrain for remembering the details I did not already know. The combination worked, but it also taught me something I did not expect going in. Flashcards can save an enormous amount of time, but only if you are selective about what you turn into a card.

This is the honest version of how it went, including the part in the middle where my own deck started working against me.

My process for the SIE

I started with Achievable's SIE course and worked through it section by section. As I read, I took screenshots of anything I did not know or believed I would need to recall on exam day. Definitions, thresholds, product characteristics, regulatory timelines, the distinctions between account types that all sound alike until you have to pick one under time pressure.

I uploaded those screenshots directly into Memgrain and asked it to generate a specific number of flashcards. Within seconds the cards were ready.

That speed mattered more than I expected. Writing a question and answer by hand for every rule, formula, product, and regulatory detail is slow work, and it is slow in a way that feels productive without actually being productive. Formatting a card is not learning. Deciding what deserves a card is learning. Handing the formatting to a tool let me spend my attention on the part that actually moves the needle.

At the peak of my SIE preparation I was reviewing around eighty flashcards a day, sometimes more. After roughly two weeks of studying, I passed.

What changed during the Series 63

After the SIE I moved on to the Series 63 and started the same way, creating cards for unfamiliar material as I encountered it. By the middle of that preparation, though, the deck had started to feel like a second job.

The problem was not that the cards were ineffective. The problem was that I was turning too much information into cards. Every new day added new reviews on top of the reviews I already owed from the SIE, and the daily queue grew faster than I could work it down. Completing reviews began to feel like an obligation instead of a tool, and I was spending time trying to hold every individual fact in memory instead of practicing how those facts would actually show up on the exam.

So I changed strategy. Instead of generating cards for everything, I shifted most of my time to practice exams in Achievable. Practice questions surfaced my recurring weak spots and got me comfortable with the wording and structure of the real test, which is its own skill. I kept reviewing my existing SIE cards, but I became far more selective about what earned a new one.

That single change made the rest of the Series 63 preparation manageable.

How spaced repetition paid off on the Series 7

The Series 7 covers considerably more material than the SIE, but a large portion of it overlaps. Because I had kept reviewing my SIE cards on a spaced schedule the whole time, that foundational material was still there when I needed it.

I did not have to reread and relearn what I had already studied. I could point my time at the parts that were genuinely new: options strategies, municipal securities detail, the deeper suitability and taxation content, and whatever the practice questions kept catching me on.

This is the part of spaced repetition that gets undersold. People talk about it as a way to pass a single exam. The real return is that it lowers the cost of the next exam. Information I would otherwise have lost within a month was still accessible when I sat down to prepare for a harder test that assumed I knew it.

If you are stacking licensing exams, or moving from one course to a related one, the compounding effect is worth more than any individual study session.

Where the AI generation actually saved time

Building a good deck by hand normally means several separate steps. Identify the important material. Write a clear question. Write an answer that is precise without being a paragraph. Decide whether the card is trying to hold too much. Then type all of it into a study system.

With Memgrain the loop collapsed to: screenshot, upload, say roughly how many cards I wanted, review the output. Seconds instead of an evening.

The second-order benefit was timing. Because card creation was nearly free, I could make cards while I was studying, at the exact moment I noticed a gap, rather than setting material aside for a separate deck-building session that I would inevitably do badly at eleven at night. The context was still in my head when the card was created, and that made the cards better.

Where I went wrong

My biggest mistake was assuming that because creating flashcards was easy, I should create flashcards for everything.

Low friction has a hidden cost. When making a card takes three seconds, there is no natural filter stopping you from saving every fact that might conceivably appear on a test. The friction of hand-writing cards used to do that filtering for you, badly but for free. Remove it and you have to supply the judgment yourself.

Eventually the review workload gets large enough that you start resenting it, and a study habit you resent is a study habit you abandon.

What actually deserves a card

• Definitions and precise terminology you cannot paraphrase your way through
• Formulas and the conditions under which each one applies
• Numeric thresholds, limits, and dollar amounts
• Timelines and deadlines, especially regulatory filing windows
• Distinctions between concepts that sound nearly identical
• Anything you have missed on a practice exam more than once

What does not

• Explanations you understand and can reconstruct from principles
• Broad conceptual material better tested through practice questions
• Context and background that supports a fact without being the fact
• Anything you added because it appeared on the page, not because you missed it

Flashcards and practice exams do different jobs. Flashcards preserve specific information over time. Practice questions teach you to apply it, recognize how a concept gets tested, and make a decision under exam conditions with four plausible-looking answers in front of you. I needed both, and when I let one crowd out the other, my preparation got worse.

"But you have a PhD in finance"

A fair response to this story is that I already had a PhD in finance, so these exams should have been trivial. That was not my experience.

A doctorate gives you deep expertise in a very narrow area. My research background did not prepare me for the breadth of products, regulations, account types, tax treatment, suitability standards, options strategies, and industry practice that licensing exams cover. In a few areas my background helped. In many others I was learning the material for the first time, the same as anyone else.

Licensing exams reward breadth. Doctoral research rewards specialization. They are different forms of knowledge, and one does not substitute for the other.

What I would do differently

I would keep the same three-part system: course material to build understanding, flashcards to retain the details, and practice questions to learn how to apply them. The thing I would change is how selectively I made cards, and I would base that decision on how much time I actually had.

Over a long study horizon, turning nearly every unfamiliar sentence into a card works fine. Spaced repetition distributes those reviews across weeks, and no single day gets crushed. You end up with broad, durable knowledge and a manageable queue.

My problem was a compressed schedule across three exams. I kept adding new material faster than the spacing algorithm could spread the reviews out. The backlog was not an algorithm failure, it was an input failure.

A rough rule I would use next time

• Two to three weeks per exam: keep the deck tight. Definitions, formulas, thresholds, timelines, easily confused pairs, and repeated misses only. Everything else goes to practice questions.
• Two to three months: go comprehensive. Card most unfamiliar material and let the spacing do its work.
• Either way: set a ceiling on daily reviews you know you will actually complete, and stop adding when you hit it consistently.

The takeaway

Memgrain let me generate study material in seconds and helped me carry knowledge from one exam into the next. My biggest lesson was not that fewer flashcards are always better. It is that the right number of flashcards depends on how much material you need to learn and how much time you have to learn it.

Making cards easy to create is enormously valuable. The next thing worth building is help matching the size and depth of a deck to your goal, your timeline, and the review load you can realistically sustain.