Can Wappler do this?

Wappler General Need Help
1

So…everyone said to ask my questions regarding what I need for my site, so here I go.
Note: I fully understand that this is a very, very complex website with very complex functionalities. I, no doubt, will have to hire some help with certain things as well as use 3rd-party companies for compute. I just want to build a bare-bones site in order to seek funding. That way, I can pay for building the very complex functionalities.
In the end, I’ll give you a real-world, very simple example that’s like what I’m trying to do.
Thanks in advance for the help.
My website needs the following functionalities:
General:
The site is to be based on several large databases of both structured and unstructured data.

  • The site needs to search through all of the user-login-based profiles (that is, if one person has only one profile and another has four profiles, the system would have to search through all five profiles). As the search criteria is entered, the system will need to create a unique algorithm and to search the site. It will then use a vector comparison (blanks are nulls) for “fit” (perhaps using Euclidean or Manhattan distance) to determine which profiles best “fit” the searcher’s unique algorithm. Once fit is determined, the system displays the profiles with the best fit. Naturally, if the unique algorithm is too specific, there would be a lower number of resultant profiles and vice versa.
  • There would, of course, need to be other types of other data analysis available – charts, graphs, et cetera.
  • The ability to post site-wide money-generating advertisements and videos.
  • A personalization system to create user-tailored feeds based on the above-mentioned advertisements and/or videos.
  • The ability to display icons next to the profiles in the search results.
    Type 1 Users:
    The ability to * handle 200 million “type 1” user logins. (Dream big right?) Note: I’m made up the “type 1” name to differentiate between another type of user.
  • The ability for each of these 200 million users to fill out forms of structured data, thus creating a profile. Each user should be able to create multiple profiles that are “attached” to their login.
  • In each profile, the user should be able to “attach” pieces of structured data from a database to their profile. The way I envision this is that they have a search bar, type in the piece of data they wish to attach, and (maybe) click to highlight that piece of data. They can do this as many times as they need to attach multiple pieces of data to their profile.
  • Next to each attached piece of structured data that has been attached to their profile, they should be able to enter a number.
  • On each user’s profile, there are free-form NoSQL fields, date fields, radio buttons, et cetera.
  • Each user should be able to post account-related documents, videos, and pictures.
  • Each user should be able to send messages to other users (of any type).
  • Each user should be able to engage in live chat operations.
    Type 2 Users:
  • The ability to handle an additional 90 million “type 2” user logins (for a total of 290 million site users).
  • The ability for each of these 90 million users to fill out forms of structured data to perform research and other database-based data analysis.
  • The ability to create multiple tab home pages featuring text, documents, videos, and pictures.
  • Each user should be able to send messages to other users (of any type) – both to specific individuals as well as to grouped individuals.
  • Each user should be able to engage in live chat operations.

Real world example:
Tim is an amazing track and cross-country runner. He wants to find someone to train with in his large city.
He goes on the site and posts the following:
Tim
Profile 1 Track Race Times
Solar Sprint Challenge Time: 59 minutes
Storm Chaser Relay Time: 69 minutes
Chronos Cup Time: 75 minutes
Geo Dash: 34 minutes
Profile 2 Trail Race Times
Rock Creek Run: 107 minutes
Rough and Tumble Run: 138 minutes
Tim finishes entering his times, and then enters other information such as his zip code to find likely runners near him.

The following runners are found by the system:
Sonya **
Profile 1 Track Race Times
Solar Sprint Challenge Time: 62 minutes
Chronos Cup Time: 90 minutes
Geo Dash: 44 minutes
Barney
Profile 1 Track Race Times
Storm Chaser Relay Time: 71 minutes
Chronos Cup Time: 74 minutes
Profile 2 Trail Race Times
Rock Creek Run: 108 minutes
Profile 3 Ultra Marathons Race Times
Ultra Winner Marathon: 240 minutes
Dog Days Ultra: 264 minutes
Jack
Profile 1 Track Race Times
Solar Sprint Challenge Time: 58 minutes
Chronos Cup Time: 74 minutes
Geo Dash: 35 minutes
Profile 2 Trail Race Times
Rock Creek Run: 105 minutes
Profile 3 Ultra Marathons Race Times
Ultra Winner Marathon: 238 minutes
Dog Days Ultra: 262 minutes
Ragged & Jagged Ultra: 247 minutes

Used Google Gemini for this:
To find Tim's best training partner, we can design an algorithm that calculates the Mean Absolute Percentage Fit based exclusively on the races both Tim and a potential partner have in common.
The Algorithm
Identify Shared Races: Filter out any races where either Tim or the other runner has a null value.
Calculate Individual Race Fit: For each shared race, determine the pacing similarity using the formula:
"Race Fit"=1-|"Runner Time" -"Tim's Time" |/"Tim's Time"
Calculate Overall Percentage Fit: Average the individual race fits and multiply by 100 to get a percentage.
The Runner Data Matrix
Race Name Tim Sonya Barney Jack
Solar Sprint 59 min 62 min null 58 min
Storm Chaser 69 min null 71 min null
Chronos Cup 75 min 90 min 74 min 74 min
Geo Dash 34 min 44 min null 35 min
Rock Creek Run 107 min null 108 min 105 min
Rough & Tumble 138 min null null null
Ultra Winner null null 240 min 238 min
Dog Days Ultra null null 264 min 262 min
Ragged & Jagged null null null 247 min
The Math & Match Calculations

  1. Sonya
    Shared Races: Solar Sprint, Chronos Cup, Geo Dash
    Math:
    Solar Sprint: 1-|62-59|/59=1-0.0508=94.92%
    Chronos Cup: 1-|90-75|/75=1-0.2000=80.00%
    Geo Dash: 1-|44-34|/34=1-0.2941=70.59%
    Average Fit: (94.92%+80.00%+70.59%)/3= 81.83%
  2. Barney
    Shared Races: Storm Chaser, Chronos Cup, Rock Creek Run
    Math:
    Storm Chaser: 1-|71-69|/69=1-0.0290=97.10%
    Chronos Cup: 1-|74-75|/75=1-0.0133=98.67%
    Rock Creek Run: 1-|108-107|/107=1-0.0093=99.07%
    Average Fit: (97.10%+98.67%+99.07%)/3= 98.28%
  3. Jack
    Shared Races: Solar Sprint, Chronos Cup, Geo Dash, Rock Creek Run
    Math:
    Solar Sprint: 1-|58-59|/59=1-0.0169=98.31%
    Chronos Cup: 1-|74-75|/75=1-0.0133=98.67%
    Geo Dash: 1-|35-34|/34=1-0.0294=97.06%
    Rock Creek Run: 1-|105-107|/107=1-0.0187=98.13%
    Average Fit: (98.31%+98.67%+97.06%+98.13%)/4= 98.04%
    Conclusion
    Tim's Best Training Partner: Barney (98.28% Fit)
    While Jack is an incredibly close second at 98.04% and actually shares more races in common with Tim (4 races vs. Barney's 3), Barney's specific running times match Tim's actual pacing slightly better overall.
2

It's not about if Wappler can build it or handle 200 million logins (which obviously you're not going to be using MySQL for that), its if you can build it. So simply put it is a 'dating website' for athletes? When you bare bones it that is what it is essentially. Yes you can build this in Wappler without issue. Doesn't matter how many databases it is but at that scale you're probably looking towards Oracle. Which currently is not within Wappler's remit. Aside from that if you can think it Wappler can build it. But Wappler can not resolve the scaling issues. Now to be honest I'm not sure if you are being serious with those numbers? Where are you getting them from? As for the rest they are just queries... Hate to be blunt but those numbers really don't make any sense.

Now I'm not going to waste time doing the math so I asked an LLM.

Scenario 1: Oracle Autonomous Database (Cloud)

Assume:

  • 200 million application logins/month
  • Each login generates 5 SQL queries
  • Total: 1 billion SQL queries/month
  • Database size: 1–5 TB
  • Moderate CPU requirements (8–16 OCPUs)

Typical monthly costs:

  • Database compute: $4,000–$12,000
  • Storage: $200–$800
  • Networking/backups: $500–$2,000

Estimated total: $5,000–$15,000/month

Scenario 2: Enterprise Oracle Database

If you're running Oracle Enterprise Edition yourself:

Example:

  • 2 servers
  • 32 CPU cores total
  • Enterprise Edition licensing
  • Annual support

The effective monthly cost can be:

  • License amortization: $20,000–$60,000/month
  • Support: $5,000–$15,000/month
  • Infrastructure: $2,000–$10,000/month

Estimated total: $30,000–$85,000/month

Scenario 3: Large Enterprise

Suppose:

  • 200 million authenticated users
  • Several billion SQL statements
  • High availability
  • Disaster recovery
  • Multiple database clusters

Monthly costs could be:

Component Monthly Cost
Oracle licensing $40k–150k
Cloud compute $20k–80k
Storage $2k–15k
Networking $5k–25k
Support $10k–30k

Total: $75,000–$300,000+/month

Important point

200 million queries is not actually a large database workload.

For perspective:

  • A busy banking system can execute billions of queries per day.
  • Large e-commerce sites routinely process tens of thousands of queries per second.

A billion SQL queries over a month averages only about:

  • 33 million queries/day
  • 1.39 million/hour
  • 386 queries/second

Modern Oracle databases can comfortably handle several hundred to several thousand queries per second on modest hardware, assuming the queries are reasonably optimized.

A practical estimate

If you built a SaaS application with:

  • 200 million logins/month
  • Around 1 billion SQL queries/month
  • 2–5 TB of data
  • Moderate complexity

You might expect:

  • Oracle Cloud: approximately $8,000–$20,000/month
  • Self-managed Oracle Enterprise: approximately $40,000–$100,000/month

The actual cost depends much more on required CPU capacity, storage, high availability, disaster recovery, and Oracle licensing model than on the raw number of queries.

So as you can see it is quite hard to take your inquiry seriously... Why are you considering Wappler for such an operation? That part makes no sense. I don't want anyone to waste their time if you are not serious... Once again apologies for being blunt and a little negative but not the average request for guidance we see here.

3

Hi,
No problem being blunt. I appreciate it. I have a thick skin.

Actually, the numbers are spot on based on what I'm building. In fact, they're a bit conservative.
Cheese - tell me more: When you say, 'Oracle," are you talking about this: Visual Builder for Web and Mobile Apps| Oracle

Are you saying that, if I'm going to build a site like this, I should use Oracle's tool?

Thanks,
Bunny

4

We all need that in this game Bunny. It really helps.

Any real-world examples you can point at?

No I'm talking about your database needs for such numbers. See above.

Not at all but with the numbers you are speaking of you'll need more than a $15.00 a month Digital Ocean database cluster (for example)...

5

OK. See...this is EXACTLY the stuff I need to know. Thanks.

6

@Cheese Hey Cheesey....how much do you think it will cost to build this thing?

7

Bit difficult to answer that without a full breakdown of what you want and how quick you want it... Front-end wise not as much as the back-end/hosting. About as much as I can tell you in that regard. Too much work on to consider it further than that I'm afraid Bunny. I wish you all the best with it though.

8

I was looking to hire you.

9

I'm incredibly busy right now. However feel free to email me info(AT)mrcheese(DOT)co(DOT)uk

10

Based on your description, the complexity of the task has less to do with Wappler (or any other development tool) and more to do with how you structure the database and backend.

As for Wappler, creating any forms, designs, icons, and so on is no problem. You mention 300 million users. It’s not clear if that’s the total or concurrent users, but that probably doesn’t matter much :slight_smile: :slight_smile: :slight_smile: with numbers like that. I work with Node.js in Wappler. So, for anything related to high load and scalability, you can read up on Node.js. As I understand it, high load and scalability are supported, but I personally haven’t implemented them yet.

As I understand it, you have a lot of algorithms and databases. None of that applies to Wappler. You can build a simple API on Wappler, but only a very simple one. Everything else needs to be done differently. I build mine in Python—it’s convenient and scales beautifully. Databases are a separate topic altogether; I use PostgreSQL.

What will Wappler offer you at the start? I think simplicity and speed—especially with AI as an assistant.

Later, once you secure funding, you can decide whether Wappler is enough for you or not. But in any case, it seems to me that most of your work won’t be related to Wappler (or any other similar development tool). Good luck with your ambitious project!

11

Regarding your huge user numbers: It probably doesn’t make sense to build a system to handle that many users right from the start. It would be unreasonably expensive in the initial phase. But if we’re talking about PostgreSQL, this database is perfectly capable of handling tens of thousands of queries per second. If necessary, you can distribute the load across multiple servers. But you don’t need that right now—you’ll get bogged down in optimization before you’ve even released the first version of the product :slight_smile: :slight_smile: It’s also impossible to accurately estimate the load without knowing what the queries will be like. If it’s an INSERT with three fields, that’s one thing, but if it’s select queries with dozens of JOINs, that’s a whole different story. Don’t get hung up on millions of users, unless you really don’t have any in your database :wink:

12

Start small first. Wait with things like the custom feeds, live chat etc. First focus on the user management and profile pages. I would suggest to start with Postgres as database. Postgres does have json fields which is ideal for storing unstructured data. Also postgres is extendable, with pgvector you can store vectors and do similarity search. I would suggest using NodeJS and also directly start with using Redis as it allows you to scale easier later.

For starting with the project Wappler is a good choice. With the requirements you have, you will probably need to build a custom backend later, there is no editor that can build that for you.

3 Likes
13

Thanks so much for your input! :slight_smile: :smiling_face_with_three_hearts:

14

Thanks for your input. :heart_eyes:

15

OK. Thanks very much for your input. :face_blowing_a_kiss: