Al Pudgino

accelerate

@mtbitcoin @AaronTeng @sssionggg @FoxyPenguinApe @PenguAsia need help with my solscan account - is it possible to send you a dm

@AaronTeng @sssionggg @FoxyPenguinApe @PenguAsia One of the most hard working folks 👍, Aaron is “everywhere”

again would love to thank @AaronTeng and @FoxyPenguinApe from @PenguAsia for the tremendous help on catlumpurr! ❤️❤️❤️

@ExitLiqCapital @GlobalWatchClub my jigga

@GlobalWatchClub @AlPudgino u on that kevin hart drip drip my boi

Kevin Hart wearing his Audemars Piguet Royal Oak Chronograph in White Ceramic

we got bigger fish to fry, brotatochip

Minnesota’s economy doesn’t work without child care providers – so we’re giving them a raise.

An Ad for the Quality Learing center 👀😂

@TommyOKid @nickshirleyy Drop the N, it’s cleaner

me after becoming a billionaire from a so called blockchain that no one uses

retard maxxing season

@crispheaney wisdom

The easiest way to get rich is expanding your time horizon The easiest way to expand your time horizon is to not care about getting rich

switching up on my day ones, now I'm popping on the streets but alone when the day’s done

my bro getting big

today is a big day, finally hit 225 on bench @GravitusApp @AlPudgino

The bull market has begun.

PRESENTING: MUSE FILMS My new production company. MELANIA, the film, exclusively in theaters worldwide on January 30th, 2026.

When I have a disagreement with a girl now, I export my entire chat history with her into AI and ask it to analyze the conversation, then paste the results to her There is absolutely nothing she can do; it's a brutal mog I then tell her to contact me after she has spoken to AI

@Lempheter @raadlabs based

@JustJake @raychen @Railway Thank you

@AlPudgino @raychen @Railway Hi Dan In a meeting talking about this right now. Broad strokes agree

@raychen @Railway @JustJake Sending a copy of my email here, since its been a few hours and I still have not gotten a response: Hi Jake and Railway team: I’m Dan (Railway Pro customer, workspace c03192e1-1918-4bbe-bde5-35483f7a16dc) I’m including Railway’s investors here (Erica Brescia & Jordan Segall at Redpoint, John Vrionis at Unusual Ventures, Lachy Groom, and Alumni Ventures) because what’s happening to our account is serious and, if not fixed, reflects directly on the way Railway treats paying customers. Summary We are on the Pro plan, running ~30–40 replicas at 32 GB RAM per replica, exactly as described in Railway’s own pricing and docs for Pro (“up to 32 GB RAM / 32 vCPU per service” and “each replica gets the full resources of your plan”). After a series of discussions about potentially upgrading to 64 GB and/or Enterprise, our account has now been unilaterally downgraded from 32 GB to 16 GB RAM across the board, without notice, without our consent, and while we are actively paying for usage at the Pro limits. When we requested a simple revert back to our original 32 GB configuration, our tickets were closed as “duplicates”, one of the threads is only visible to our developer and not to me (the account owner), and we’re now being told that we must sign a $2,000/month pre-commit just to restore the resources we were already using and paying for. These changes were done mid-week / over the weekend and directly impact core indexing services for our product. We are not asking for anything exotic; we are asking to be restored to the published Pro-plan configuration we have been running and paying for. Timeline & facts (with references) Existing paid configuration (what we had before this week) We have been running 30–40 replicas at 32 GB RAM on the Pro plan for some time. This is exactly in line with Railway’s pricing & docs for Pro, which say: Pro plan: up to 32 GB RAM / 32 vCPU per service. “If you’re on the Pro plan, each replica gets 32 vCPU and 32 GB RAM. Deploying more replicas multiplies that capacity.” Conversation with Rahul about upgrading to 64 GB (Discord, Nov 13) I asked about upgrading to 64 GB of RAM. Rahul explained that 64 GB is only available on Enterprise, with a $2,000/month minimum spend on an annual commit, and described higher CPU/RAM/storage/PID/IOPS limits and SSO/audit logs. I pushed back that we’re an early-stage, bootstrapped startup and cannot commit to $2,000/month for a year right now. Rahul confirmed that to get 64 GB we’d need Enterprise, but we discussed the need very clearly and he stated he wouldn’t “upgrade and then deny that”, and that he would coordinate with infra to make sure it was approved before anything was finalized. Google Meet with Angelo and trial of 64 GB configuration (Thursday night) We then spoke with Angelo and the infra team on a Google Meet (Thursday). They proposed a compromise: a smaller number of machines at 64 GB (fewer replicas) instead of 30–40 at 32 GB, as a short-term path. We tried this configuration. It did not work for our workload. I informed the team quickly that we needed to revert and, in the meantime, we’d continue to refactor to reduce RAM usage. Our request: revert to original 32 GB Pro configuration After confirming the 64 GB test wouldn’t work for us, I opened support threads and posted in Discord asking to revert to our original configuration: 32 GB per replica, ~30–40 replicas, on Pro. Our dev Mike also responded in the Central Station ticket explaining that, prior to the 64 GB request, we had 30–40 instances at 32 GB with no pre-commit and no issues. Support behavior & closed tickets My revert-request thread in Central Station was closed and I was redirected to a thread that: I cannot see as the account owner (only my dev can see it because he originally opened it). I was then told not to open “duplicate” threads when I tried to re-open the conversation, even though I was simply trying to undo a change that Railway had made. In Discord, when I asked for help reverting the changes and tagged the team, I was told not to tag them and to use “formal support channels” – the same channels where my revert requests were being closed. All of this is happening while our workload is in a degraded state relative to where it was before we ever asked about 64 GB. Unilateral downgrade from 32 GB to 16 GB & new pre-commit requirement Today, my dev informed me that Railway has reduced our RAM limits from 32 GB to 16 GB across the board. This is a downgrade below the published Pro plan limit and below our historical configuration, and it was done without prior notice and without any request from us. He was told that Railway now requires a $2,000/month pre-commit just to return us to our original 32 GB configuration that we had been running and paying for all along. Why this is so problematic It contradicts your own published Pro-plan description. Your docs and pricing page clearly advertise that Pro supports up to 32 GB RAM per service/replica, and that every replica gets those full limits. We were operating exactly within those published limits and paying usage-based charges for it. You are changing critical resources mid-stream, without notice, on a paying account. Going from 32 GB to 16 GB RAM across our services – during the week / weekend – could cause production workloads to run out of memory. If we had mission-critical processes, this would have been catastrophic. We are being told to pre-commit to Enterprise-level spend just to get back what we already had. We’re not insisting on 64 GB right now. We explicitly said we’re willing to: Stay on Pro Keep paying for our actual usage Continue refactoring our workload to be more efficient We are simply asking to be restored to the configuration we had and that your own plan promises. The support experience has shut down our ability to resolve this through normal channels. Tickets closed as “duplicates,” a key thread that the account owner can’t even see, and admonishments for tagging the team in Discord – all while the underlying resource change remains unresolved – do not line up with “priority support” for Pro customers. What we are asking for I want to be very explicit and reasonable about the remedy: Immediate restoration of our original Pro configuration: 32 GB RAM per replica in the region where we were previously running, The ability to run 30–40 replicas, as we have historically, On the Pro plan, billed according to your normal usage-based model. Written confirmation that Railway will not: Reduce our per-service RAM below the Pro plan’s published limits, or Cut our RAM in half again, without explicit notice and our agreement, or at minimum with enough advance notice for us to migrate safely. We have invested significant engineering time and business risk into building on Railway. Up until this incident, we’ve been advocates of the platform. But unilateral downgrades of paid capacity, combined with tickets being closed while we ask for a revert, are not acceptable for a production-oriented, paying customer. I’m happy to jump on a call early this week with whoever on the leadership/infra side is appropriate (Jake, Angelo, Rahul, etc.) along with our dev, to resolve this quickly. What we can’t accept is remaining stuck at 16 GB with no path back to the Pro-plan configuration we were originally sold, unless we sign up for an Enterprise pre-commit. Thank you for your time and attention to this.