What is the best long term strategy for a company that wants to develop their own application(s), but is not an software development company, but a logistic supply chain company with own terminal and transporting company.
Develop software ai suported in c#/.net or ai suported in lowcode platform
I’ve been rebuilding part of our email workflow lately and honestly the more I compare them, the more I lean toward using a dedicated tool like Stripo email builder instead of Mailchimp’s built-in editor. Mailchimp is okay for super basic newsletters, but once emails become more complex the editing starts feeling restrictive fast. Reusable modules, responsive behavior, cleaner exports, and keeping templates consistent across campaigns all seem easier with Stripo email builder. Biggest thing for me is maintainability. I’d rather spend time optimizing campaigns than constantly fixing spacing/rendering issues every time someone edits a template inside Mailchimp. If anyone else ended up separating email design from the sending platform for the same reason.
Hey all! I’ve recently been getting deeper into workflow and IT process automation and wanted to start documenting more of the things I build and learn along the way.
I’ve been working with tools like Workato, n8n, Okta Workflows, Power Automate, Slack Workflows, and Claude, and I’m looking forward to experimenting with even more tools and integrations.
I’ll mainly be sharing workflow ideas, automations, and process improvement projects as I continue learning and building in this space.
Would also love to connect with others interested in automation or low-code workflows. I’ll be sharing some of this on IG too: @sheautomated
Been poking around with some of the Gemini-based workflow stuff lately, specifically looking at whether it can, slot into our lead routing and enrichment pipelines without needing a dev involved every time something breaks. BuildShip and AppSheet both look promising on paper, and the Gmail trigger stuff in Workspace is genuinely handy for some of our ops use cases. But every time I get into the actual config, there's always some JSON handling or rate, limit quirk that makes me wonder if the 'no-code' label is doing a bit of heavy lifting. The community vibe I've seen elsewhere is that Gemini is decent for drafting and summarisation, but gets inconsistent when you need it to reliably handle logic across multiple steps or files. For anyone actually running Gemini integrations in a sales or revenue ops context, how's the reliability holding up in production? I keep seeing n8n pop up as the go-to for people who want more, control, but curious if anyone's found a setup that genuinely stays stable without constant babysitting.
Running a solo operation and trying to figure out which tools actually pull their weight, for scheduling, social media, and basic customer support without needing a whole team to maintain them.
Zapier is the obvious starting point, setup is dead simple and the native integrations are, hard to beat, but the pricing gets painful fast once you're running any real volume. n8n on the other hand gives you way more flexibility and the self-hosted option keeps costs, low, though the learning curve is noticeably steeper if you're not comfortable with JSON and node configs.
I've also poked around Make and Latenode for the workflow side, and both sit somewhere in, the middle on complexity vs cost, which might actually fit a solopreneur better than either extreme.
Priorities for me are: easy enough to set up solo, hard to accidentally break, affordable at low-to-mid volume, and decent enough at handling support-adjacent tasks without babysitting.
Which of these would you actually trust to run your business ops day-to-day with minimal maintenance?
Running SEO ops for a mid-size SaaS and we're finally replacing a patchwork of Zapier, zaps with something that can actually handle multi-step AI agent workflows, not just linear triggers.
Make.com has been solid for visual logic and the scenario debugger is genuinely useful, but the per-operation pricing gets ugly fast once you're running AI classification steps at volume. n8n self-hosted is cheaper long-term and the node flexibility is hard to beat, but our ops team isn't technical enough to maintain it without me babysitting every update.
I also poked around Latenode since it prices on execution time rather than per-op, which changes the math considerably for AI-heavy flows.
The factors I'm weighing, roughly in order: cost at scale, how much the non-dev team can own, day-to-day, quality of AI model access without separate subscriptions, and error visibility when an agent step fails mid-run.
The part I'm least sure about is the infrastructure side, specifically whether Make or n8n handles agent steps, that need to write back to external APIs reliably under load, or if that's where both start to crack.