Selim Ekizoglu

[Selim Ekizoglu @selimekizoglu]

@BraedendotTECH with `codex --yolo` you can even avoid the "enter" burden :)

[Selim Ekizoglu @selimekizoglu]

@danielstuski @launchable_ @better_reviews go back to work but keep working on my side project after 5pm

[Selim Ekizoglu @selimekizoglu]

@NicolasLecocqHQ what will you be doing differently this time? any learnings worth mentioning?

[Selim Ekizoglu @selimekizoglu]

@oalpay3 any agents working on distribution actively?

[Selim Ekizoglu @selimekizoglu]

@cnye36 Do you end up using Storage, Delay, Code, dedupe, cooldowns, etc often?

Mind sharing few such workflows?

I’ll check whether the stateful part fits the webhook rule engine I’m building and what it could potentially save or simplify.

[Selim Ekizoglu @selimekizoglu]

@omarvvvr Cancel subscription

[Selim Ekizoglu @selimekizoglu]

@marconoiz Hallucinate

[Selim Ekizoglu @selimekizoglu]

Have a Zapier/Make/n8n workflow where webhooks need state?

Storage, Delay, Code, dedupe, cooldowns, “wait for X unless Y happens,” etc.

Send it to me.

I’ll check whether the stateful part fits the webhook rule engine I’m building and what it could potentially save or simplify.

[Selim Ekizoglu @selimekizoglu]

@berkulsoy_tr @denizoktar Yep, makes sense. I should’ve scoped it better.

I’m mainly curious about small SaaS/ops teams - a few people, no dedicated workflow/platform ownership - and how they solve stateful webhook rules today.

[Selim Ekizoglu @selimekizoglu]

Have you ever needed Zapier, Make, or n8n to trigger based on a sequence of events rather than a single event?

Example:
A happened → B happened within 7 days → C did not happen → then alert someone.

How did you solve it - storage tables, delays, filters, custom code?

[Selim Ekizoglu @selimekizoglu]

Zapier is great for “event → action.”

But what about workflows that need memory across events?

"Same customer hits multiple risk signals across billing/support/CRM within 7 days → alert CS once."

Delay handles time. But where do you keep counts, joins, dedupe, and expiry?

[Selim Ekizoglu @selimekizoglu]

@cyber_gilgamesh Thanks — makes sense for a linear sequence.

What if it’s per account + count/window based?

e.g. same account has 3 failed payments, 2 support tickets created, and a downgrade request within 7 days.

Would you still use sleep/if/route, or store state somewhere?

[Selim Ekizoglu @selimekizoglu]

Question for no-code automation builders:

How do you handle automations that need to count and match events across multiple tools?

e.g: Customer has 3+ failed payments in Stripe, 2+ tickets in Zendesk, and tagged as a churn-risk in HubSpot within 7 days → alert CS.

[Selim Ekizoglu @selimekizoglu]

Question for no-code automation builders:

How do you handle workflows that depend on events across multiple SaaS tools over time?

e.g. A happens in Stripe, B happens in HubSpot, C doesn’t happen within 3 days → do something.

Do you use storage tables, Airtable, delays, etc?

[Selim Ekizoglu @selimekizoglu]

@cnye36 Makes sense. The mapping step seems like the real unlock.

Do you ever run into workflows where the hard part isn’t the steps, but the state/timing around them?

Like “this happened, then this other thing happened later, but only if something else didn’t happen in between.”

[Selim Ekizoglu @selimekizoglu]

Zapier folks - how do you handle Zaps that need memory?

Ex: 3 events for same customer in 24h, or A happened but B never followed up.

Storage? Delay? Paths? Code? Separate DB?

[Selim Ekizoglu @selimekizoglu]

Curious how you handle scenarios that need state.

Not just “webhook → run scenario,” but:

3 events for same customer in 24h
A happened, then B happened later (or B never followed up)
Stripe + HubSpot + app events joined by customer_id

Data stores? Routers? Sleep? Custom API?

[Selim Ekizoglu @selimekizoglu]

n8n folks,

when a webhook workflow needs state, where do you put it?

Not just “event in → run workflow,” but rules like:

3 events for same customer in 24h
A happened but B never happened
Stripe + PostHog + HubSpot events joined by customer_id

Wait nodes? Data stores? DB?

[Selim Ekizoglu @selimekizoglu]

@berkulsoy_tr @denizoktar i.e. teams without engineering ownership for this layer — they can wire webhooks/APIs, but don’t have someone maintaining a BPMN platform + connector catalog. Would you still point them to BPMN?

[Selim Ekizoglu @selimekizoglu]

@berkulsoy_tr @denizoktar Makes sense. I guess I’m probing the boundary: for a smaller SaaS-ops team with just webhooks/APIs and a few grouped temporal rules, would you still adopt BPMN, or is that too much platform?

[Selim Ekizoglu @selimekizoglu]

@denizoktar :) iyilik valla. Yan proje kovaliyorum :)