Hi HN,

I've been exploring a different problem space from AI agents and LLM frameworks.

Many tools focus on prompts, orchestration, and model capabilities, but production systems often fail because of runtime issues: retries, state loss, failed jobs, crashes, and recovery.

I built ALGOgent Runtime SDK, a lightweight Python toolkit that adds:

- Retry engine - Backoff strategies - State persistence - Checkpoints - Event bus - Logging - Metrics

The goal is simple: make automation scripts and AI workflows more resilient without requiring cloud services, GPUs, or heavy dependencies.

Project page: https://stateflow-dev.github.io/algogent/

I'd appreciate feedback on the concept and whether this is a problem developers actually experience in production.

I built this after my AI agent kept crashing in production and losing state. Happy to answer any questions about the architecture or benchmarks!

"The surge in math deficiencies after dropping the SAT highlights a systemic issue: grade inflation. Without a standardized baseline like the SAT/ACT, a 4.0 GPA from a high school with relaxed standards looks identical to a 4.0 from a highly rigorous one.

Paradoxically, removing test requirements harms underprivileged students the most. Preparing for the SAT requires a book and an internet connection. In contrast, building a competitive profile based entirely on expensive extracurriculars, sports, and elite summer camps is far more wealth-dependent. Standardized testing isn't perfect, but it's often the only objective equalizer we have."

I wasn't underprivileged but I did go to a terrible evangelical high school that had no honors or AP classes (AP bio at a place teaching creationism would've been something else...) and I think I only got in to a decent college on the strength of my SAT and ACT scores. My grades were OK (except in bio, where I refused to acknowledge young Earth creationism) but not amazing.

> My grades were OK (except in bio, where I refused to acknowledge young Earth creationism) but not amazing.

This is... Wild.

The funny thing is the Adventists seem to produce good hospitals but are still creationists. I guess it’s not a big deal how we got here if you just want to do medicine.

It's very common in US private schools.

I think this depends a lot on how you select the set of private schools you're looking at.

Please provide a reference to this.

Uh, it's been a while since I've been inside one, but I would guess it's very common in a certain strain of US private schools, not as a rule.

No it is not lol. Incredibly rare

Who gets to set the curriculum is a much bigger deal than given credit for. So many teachers complaining about the shit they have to teach. I remember one who didn't necessarily disagree but wondered why Al Gore should be the one to decide what goes into the [mandatory] documentary (in the Netherlands)

An inconvenient truth was mandatory viewing in NL?

That's what she said.

I can't read the article - do they explain why they think this is a "paradox"?

Expectation: removing standardized tests will give more opportunity to students who historically tend to do worse on those tests, like poor kids.

Reality: removing standardized tests means that universities have to put more weight on the rest of the college application, such as extracurricular activities which are often expensive and thus disadvantage poor kids.

Calling it a "paradox" is maybe a little hyperbolic, but basically it did the opposite of what they expected.

But this is relative, right? We're talking about SATs vs just relying on grades. Do poor kids do worse on SATs relative to how they do in their class grades, as compared to other kids? I kind of just figured poor kids do worse overall.

Its easier to get good grades in poor schools since regardless how you grade your teachers grade you against your peers, and poor schools have worse peers. So given same knowledge poor students have better grades, so their grades will be relatively better than their SATs.

At least in public schools were you don't have an incentive to give everyone an A, maybe private schools are different.

This doesn't exactly answer your question, but MIT added test scores back recently and wrote a blog post explaining why: https://mitadmissions.org/blogs/entry/we-are-reinstating-our...

I think the "paradox" is that you'd expect disadvantaged students to perform worse on standardized testing.

Personally, I don't think they actually believe it's paradoxical, I think the authors are just trying to be polite to those who criticize standardized testing with identity politics. Politeness can aid in persuasion, so I don't blame them.

I've been wondering with all the data that's available now couldn't admissions look at a 4.0 from HS A vs a 4.0 at HS B and then compare those to actual grades on the campus once students were in class? Assuming HS A has lower standards, they should be able to tell that a 4.0 isnt as meaningful as a 4.0 from HS B. Seems like a straightforward exercise.

I don't think it's paradoxical at all. This was the original strength of the SAT system.

And SAT as high school math exam itself I think is way too easy. They should design another test which can clearly distinguish top 1% or even 0.1%.from others

When I was in high school in California more than 20 years ago, SAT math alone was insufficient for admissions to STEM programs at mid-ranked and top-ranked universities. I was required to take the SAT Math IIC subject test, which went up to pre-calculus. We were also strongly encouraged to take calculus in high school. There are two AP Calculus exams: AB (which covers the first semester of university calculus) and BC (which covers the first two semesters).

There are already such tests. They're called International ___ Olympiad.

Yes, the scores at the top are way too bunched. A perfect score should indicate generational genius, not the 100th smartest kid your year in California.

That's not a real problem for UC admissions. They accept thousands of students every year. Anyone who scores near perfect (within the margin of error) should be admitted to at least one UC campus. If that's not happening then the problem is with the admissions criteria, not with the SAT.

> Anyone who scores near perfect (within the margin of error) should be admitted to at least one UC campus

You really don't see a problem with that? Our best students should not be "admitted to at least one UC campus". They should have their pick of the UCs and should be getting generous scholarships to entice them. There is absolutely loads of bunching near the top the distribution which makes it impossible for actually-top students to differentiate themselves.

There's about 500k kids in California each year. The 100th smartest is undoubtedly smarter than you or me.

Probably. And they deserve to be able to prove it in a way that the current tests do not allow due to their low ceiling.

Yup. SAT is just a _bad_ test. It's way too easy to get top scores there, and the whole calculator section should be removed.

SAT also makes it easy to just mechanically grind it because the problems are so simple that you can just memorize their "shapes".

We need a test with a smaller number of tasks escalating in complexity. So that the top scores are extremely rare.

> They should design another test

Why? What would the point of this be beyond bragging rights?

My kids were able to take some SAT test prep course through their school (partially funded by the PTA) and it helped a lot. They wrote a bunch of practice exams and each time their scores went up. Also, test taking itself is a skill and the more you practice it the better you get at it. If you’ve written the SAT 15 times over the past 2 years, then the 16th time won’t be as stressful and you will know strategies that work and the questions will be familiar.

If you are in a school that doesn’t have a well funded PTA, you are at a disadvantage.

You can, as of about a year ago, take official SAT practice exams for free in Google Gemini.

The person to whom I responded seemed to imply that it consists chiefly or entirely of taking practice exams. I merely wish to point out that if you want your kid to take SAT practice exams every month you can do it for free at home.

Such a "SAT test prep course" is going to involve more than just self-guided practice exams. It'll include feedback and coaching to address deficits revealed by those practice exams.

This is exactly right. Writing each practice exam only takes a few hours and this course last months. The reset of the time is filled with all the things you talked about.

Plus, for some kids writing a practice exam at home isn’t the same thing as a simulated seating with kids all around and a proctor in the room.

You seem to be arguing with someone else.

No; I'm saying "just take practice exams" isn't what we're really talking about. They are merely a part of high-end test prep offerings.

I actually took fairly involved test prep way back when (and didn't end up actually writing the SAT) and there really wasn't anything to it beyond the practice exams that I couldn't have figured out myself.

Whatever gates you put up, the wealthy can fire cannons of cash at them. You just have to pick the ones least vulnerable to cash barrages.

What is the marginal gain of expensive SAT prep? Versus just doing hundreds of mock tests out of some prep book, like SWEs grinding LeetCode?

Your analogy works against you, given that tons of professional athletes come from poverty.

Professional athletes are like people who get 1600s on the SAT; a bit of an outlier.

We're talking about the California state education system here. They do not have the option to restrict the provision of their services to a tiny elite. The concerns of "top schools" absorbs altogether too much oxygen.

IMHO, California state higher education is setup to be tiered. UC > CSU > Community Colleges. If UC is getting a lot of STEM students that need remedial math, I think something has gone wrong. Those students might be better served by getting their math needs met at a community college and transfering to UC later.

For one, why pay UC prices for remedial math? For two, community college has a lot more sections of remedial math and more experience teaching it.

If you're in a degree that doesn't need much math, taking remedial math at UC is probably fine; but all the STEM degrees want at least the full calculus series (afaik).

Remedial math for STEM students at CSU is probably in the middle. You still don't really want a lot of students in that group, when they could be better served at community college ... but CSU should also be more prepared for it.

> tons of professional athletes come from poverty

Is that actually the case?

Depends on the sport. I don’t think the Olympic equestrian competitors would be dirt poor.

Read up on Kobe Bryant or Bronny James.

Sure, those are some good counterexamples: both sons of professional athletes. And there are plenty of others.

On the other hand, we have: Allen Iverson, Larry Bird, Shaquille O'Neal, Carmelo Anthony, Michael Vick, Bo Jackson, Jackie Robinson, Babe Ruth, Mickey Mantle, Fernando Valenzuela, Albert Pujols, Jim Thorpe, ...

Oh, and LeBron James himself!

So my view is that people of both rich and poor upbringings have a good chance in the sports world these days, at least for those sports where the necessary gear is relatively cheap.

> On the other hand

Perhaps I should have instead said "is that still happening at meaningful rates".

LeBron James is an interesting example. Per wiki: > Realizing that her son would be better off in a more stable family environment, Gloria allowed him to move in with the family of Frank Walker, a local youth football coach who introduced James to basketball when he was nine years old.

and then later he went to a fancy private high school (whose wikipedia page has many notable alumni, all athletes).

So while "from poverty" may be technically accurate, I don't know if I'd count it given all his opportunities later in childhood.

Times have changed. Due to the rise of expensive youth travel club sports leagues I suspect we will see fewer poor children turn professional. There will always be a few outliers but if you don't have access to top coaching and extra competitive playing time prior to college then you're really at a disadvantage.

According to IA this is mostly a myth though.

Agreed. Professional sports are the closest institution that society has to a meritocracy. Highly competitive, public, obsessively measured and analyzed. A tenth of a second faster sprint time might be more valuable than even a top-tier socioeconomic background.

Sports is the most expensive way to get into college. Tennis is close to $1 million to get your kid into an Ivy league through tennis. Malcom Gladwell wrote about sports and colleges in his book "revenge of the tipping point". Sports is used by the wealthy to get their less academically inclined children in to top schools and some school are expanding it.

That's not the reality for most youth sports anymore. It's gotten much more competitive. Participating in school sports isn't enough. They generally can't develop the level of skill necessary to gain advantage in college admissions without paying a lot to participate in travel club teams and for private coaching. And I'm not talking just about NCAA recruited athletic scholarships but even for the sort of regular extracurricular sports activities that might give someone an advantage in college admissions.

A friend of mine brought his girl up in club soccer, he was starting to travel to different states for tournaments when she was about 8 years old. It was insanely expensive. The only thing i can think of crazier than club soccer is maybe club cheerleading. Youth sports is completely bonkers if you've never been exposed to it.

It feels like the problem are the SAT prep courses' existence then

The problem is as never the tests. It was pretending that the difference between a 600 and 625 (or whatever) really predicted anything.

It was the silly idea that with tests you could produce a fair ordering of students based on potential to succeed.

You can absolutely make a bet on who's more likely to succeed based on a 100 point difference, though. It's not absolute, but it's highly predictive. And the reason the SAT was dropped wasn't because admissions were being forced to blindly accept 620 over 610 (they never were), but so that people who scored hundreds of points below the mean could be admitted (in the pursuit of other institutional goals).

We have decades of data (test score vs grades and degree completion). They should gather it up and calculate the answers.

Flip answer: the bucket width should be 2.5 times the score improved of a prep course.

Any working system has to rely on some arbitrary rules. Drawing a line between students who scored 600 and 625 is still infinitely better than drawing it based on the decision-makers' moods.

Or, treat 600-625 as a tie, and use a lottery.

As imperfect standardized tests are, they are still more fair and less biased than using arbitrary judgement on extra curriculars

Bucket to the observed predictive power of the score, resolve ties with a lottery .

Would this be fixed buckets? I.e. would you treat 649-650 more predictive than 648-649? Presumably that wouldn't work. I'm sure there's some algorithm that could do this but it seems subtle.

Obviously, if a school has a cutoff score bucketing is easy, but with excess applicants ordering becomes necessary. I guess this sort of probabilistic score would induce an order for any given student relative to sufficiently superior or inferior applicants.... I'm now kinda curious to figure this problem out. Did not expect an algorithms problem to arise in this thread lol

And you don’t want a 100% cutoff. You need to admit some people just under the threshold since the scores are relative. You need fresh data to keep the model tuned.

Some kind of weighted lottery

Allow me to propose a model for this score-based ordering with fuzziness. (Perhaps we can call this problem probabilistic rasterization.)

The final output of an execution of the system, given a static, complete set of applicants is a particular ordering of applicants. Since lottery is involved, there are multiple acceptable orderings for a given input set. The question is to define a set of criteria to classify acceptable orderings, and a desired probability distribution of orderings, which can be satisfied by an algorithm for a maximal proportion of inputs.

For example, given a set of applicants A with score function F, we notate an ordering relation R(x,y) such that, given a limited number of seats, applicant y will be admitted before applicant x. For shorthand, x < y means R(x,y).

Possible acceptance criteria for an ordering R may include:

(1) Given some d in the codomain of F (presumably a group), FOR ALL x,y in A, if F(x) + d ≤ F(y), then x < y

Possible criteria for the distribution of orderings may include:

(1) FOR ALL x,y in A, if F(x) = F(y) then P(x < y) = P(x > y)

who uses SAT scores as "potential succeed"??

The original argument for standardized tests was to pick based on how well you would do in university (vs who your parents know).