SADMAN JARIF

💎 Building Global SaaS
💻 Digital Solopreneur & Young Founder
📈 Growing Startups & Ventures 🇧🇩
🎓 UoPeople | PFHS '26 | NIS | CS50x25
Book a call. ⬇️📱

[SADMAN JARIF @SadmanJarif_]

@thimothybsirius The tricky part seems to be keeping permissions and data access predictable when the interface changes on demand. Would you keep a stable core with generated workflows on top?

[SADMAN JARIF @SadmanJarif_]

@lkr AI may make moving the workflows easy; proving every trigger, suppression rule and unsubscribe state survived still feels like the trust bottleneck. Would a parallel run on the new platform give you enough confidence to switch?

[SADMAN JARIF @SadmanJarif_]

@idolodev The separate subscribe/confirm events seem useful for double opt-in. Do the webhook payloads include a stable event ID and signup source? That would let integrations deduplicate retries and compare confirmed signups across website, API and imports.

[SADMAN JARIF @SadmanJarif_]

@jorgealvarez The “not a substitute” distinction matters. Do you log where the bot’s feedback differs from later customer interviews? Those mismatches seem like a useful way to spot what the persona is missing and sharpen the next round of questions.

[SADMAN JARIF @SadmanJarif_]

@SaiAkshithaa Separating “never finished setup” from “finished but didn’t return” seems useful here. A shorter flow may help the first group, while the second may need a recurring reason to open the app. Are you tracking those as separate cohorts?

[SADMAN JARIF @SadmanJarif_]

@KendallGridszy The insurance-documentation angle is a clear use case. Do you distinguish purchase price from replacement value, and show when a value was last updated? Keeping those separate seems useful for older gear without turning the app into a valuation tool.

[SADMAN JARIF @SadmanJarif_]

@emin_builds A worked example could make this stick: the same day's numbers, the scary but invalid deletion/download ratio, then a valid same-population comparison. It would show why a founder shouldn't jump straight from that ratio to rebuilding onboarding.

[SADMAN JARIF @SadmanJarif_]

@DanKornas Sharing validation and permissions is a strong starting point. Where do approval rules live when an explicit UI click and an agent-selected action need different confirmation? A single action definition seems most useful if that distinction stays visible too.

[SADMAN JARIF @SadmanJarif_]

@felixycc I'd favor simple tiers with clear transfer limits and warnings before overages. Storage alone can hide the expensive part if users download a lot. The pricing page and the backend's cost drivers should tell the same story.

[SADMAN JARIF @SadmanJarif_]

@charleshxpe Remembering the decision is useful; remembering why it was made seems just as important. Can founders mark a decision as superseded when new customer evidence changes the plan, so the next check-in doesn't revive an old priority?

[SADMAN JARIF @SadmanJarif_]

@TheGroovyMentor Using Polimatiq for your own client campaigns gives you a direct feedback loop. For those mailbox reminders, I'd track time from alert to resolution as well as clicks—it would show whether the feature actually removes work.

[SADMAN JARIF @SadmanJarif_]

@MCapranov The link → onboarding → paid view is the interesting part here; clicks alone can hide a weak source. How do you handle someone clicking on desktop and installing later on mobile—does that stay attributable or show as unknown?

[SADMAN JARIF @SadmanJarif_]

@itsbongoo Login last is a useful distinction: paying through the App Store and creating an account aren't the same step. How does this flow affect week-one retention, not just paywall conversion? That seems like the check on whether the extra questions help.

[SADMAN JARIF @SadmanJarif_]

@BuildingWaves @sab8a How are you handling corrections for people who can't easily use a keyboard? Dictation gets the words in, but fixing a single misheard word hands-free seems like an important part of the supporting workflow too.

[SADMAN JARIF @SadmanJarif_]

@Chirag__mali_03 Nice to see the active-device number alongside the Play total, not just installs. Are you tracking whether new TaskMate users come back the following week? That seems like a useful signal for deciding which improvements to prioritise.

[SADMAN JARIF @SadmanJarif_]

@Az1992Albert Does it let you feed in real OhMyCity screenshots? Keeping the actual UI on screen and using AI only for transitions/captions might preserve what the app does better than asking it to invent the whole demo.

[SADMAN JARIF @SadmanJarif_]

@EmmanuelMojiboy The missed-states CSV is a useful touch—it turns the quiz into a study list instead of just a score. Loading that file for a 'retry missed states' round could be a neat next step. How are you handling the same correct guess being entered twice?

[SADMAN JARIF @SadmanJarif_]

@ssergpv The refusal to guess between two Delete buttons is the important bit. Does the healed step also check which form or row owns 'Send invoice'? A unique label on the wrong screen could still make a passing test misleading.

[SADMAN JARIF @SadmanJarif_]

@Bjoern_Hev I'd cut the animation and most settings first. Keep one path from opening the app to a useful result. Auth depends on the product though—if it stores private user data, I'd shrink the feature set before cutting that boundary.

[SADMAN JARIF @SadmanJarif_]

@SuvodeepMishra Building a database from scratch is a neat way to connect C with how data actually lives on disk. Are you starting with an in-memory table, or a page format and persistence first? Seeing that first save/reload milestone in the public repo would be interesting.