Asifuzzaman Arko

@Mikael Mikael on the tool question both work but n8n is the stronger long-term choice for client delivery, mainly because self-hosting means you're not charging clients for operation costs and you have full control over the workflows you hand off Make.com is faster to learn but as projects get more complex the scenario structure starts limiting you and the costs can creep up but I noticed you mentioned you haven't got your first client yet, that's usually the harder part and the tool choice often becomes a distraction from the actual problem which is getting someone to say yes what does your current outreach look like and what niche are you focused on

Cem Anbarcı

Apr 17 •

🧲 leads-help

how to write cold email for dental clinics

hi guys ,ı have built voice agent and automations for dentist in turkey ,before you will ask that no this market has not saturred for now in turkey ,so how can ı write the best shape of email for dentist to reach them and open the emails and how ı can reduce cost email which tool what should ı use what have you experienced on your businesses

New comment Apr 19

Asifuzzaman Arko

0 likes • Apr 18

@Cem Anbarcı since email open rates in Turkey are basically zero for cold outreach to dentists this is actually a channel-fit problem more than a copywriting problem if they don't open emails from companies they don't know, no subject line will fix that the smarter move for your market is to reach them where they are already active and attentive which for local dental clinic owners in Turkey is likely Instagram DMs, LinkedIn, or WhatsApp Business if you have any referral path into them for cold email if you still want to test it, the only thing that consistently improves open rates in cold B2B is hyper-personalization in the subject line, something that references something specific about their clinic rather than a generic hook for cost Brevo gives you 300 emails per day free which is enough to test before spending anything are you targeting dentists you've found online or do you have any warm introductions at all in the niche

Kevin Alldread

Apr 14 •

🛠️ dev-help

Is selling Voice Agents Still Viable?

Selling Voice Agents with a package of automations was my initial plan in this space for local home service business's. I have read very mixed reviews on if this is dead or still a valuable tool to be selling. Looking for advice from people actually delivering these in 2026. If they are still valuable, what platform are you using for delivery? Retell, Eleven Labs Agents etc?

New comment Apr 18

Asifuzzaman Arko

1 like • Apr 18

@Kevin Alldread voice agents are still very much viable but the mixed reviews you're reading are mostly from people who tried to sell them as a standalone product to the wrong market for local home service businesses the use case is actually strong because missed calls directly equal lost revenue and most of these owners know it, a plumber who misses a call at 9pm because no one is in the office has already lost money the platform question matters less than the integration layer, Retell and Vapi both work but where builders usually get stuck is connecting the agent output to something the business already uses like their CRM, scheduling tool, or job management platform if you can show them that the voice agent doesn't just answer calls but actually books the job into their existing system without them touching anything, that's when it stops being a tech demo and starts being something they will pay for monthly have you already built any demos for the home service niche or are you still figuring out the stack

Wiebke Heynen

Apr 18 •

🛠️ dev-help

Help: Consistent Product Photography

Hi everyone, I'm completely new to AI image tools and could really use guidance. My situation: I am about to pitch for a jewelry company. Their website has beautiful, consistent product photos — every ring is shown the same way: clean white background, same angle, same soft lighting, same subtle mirror reflection underneath. It looks high-end and uniform across all products. My problem: When a new ring arrives, I will only get a quick iPhone photo. Bad lighting, random background, wrong angle. I need to turn that iPhone snapshot into a photo that looks exactly like the other rings on their website — same background, same angle, same lighting, same reflection. Only the ring itself should change. What I've tried: I used Leonardo.ai. (simply because I have some tokens left on there, and it has different tools such as nano banana). The results aren't good — it changes the shape of the ring, messes up the engraving, and I can't get the background and reflection to look consistent. What I need help with: - Which AI tool is actually best for this? (I've heard names like ComfyUI, Flux, Midjourney, Nano Banana, but I don't know which one fits my case) - Is there a simple workflow that I can make specifically for product photos where I just upload my iPhone picture and it does the rest? Maybe a skill on GitHub for Claude Clode (I have been trying to find one...) - How do I make sure the ring itself stays exactly the same and only the background/lighting changes? - If anyone has a step-by-step workflow they'd share, I would be incredibly grateful. I have added example photos (website style + an iPhone input photo) if that helps you understand what I'm aiming for. Thank you so much — I've been stuck on this for a while and don't even know where to start. I have managed - after 100 tries, to get the signet ring into the right angle, background and lightning, but I fail to understand how to lock this so I can repeat it for other rings.

New comment Apr 24

Asifuzzaman Arko

0 likes • Apr 18

@Wiebke Heynen the challenge you're running into is a very common one with product image consistency and the root issue is that diffusion-based tools are generative by nature so they tend to reimagine rather than preserve for your exact use case the approach that works best is image-to-image with a strong reference image rather than text-to-image, tools like ComfyUI with a ControlNet setup let you lock specific visual properties like background, lighting angle, and reflection while only changing the subject the key workflow is: use your best existing product photo as a style reference, create an inpainting mask that isolates just the ring from a clean reference shot, then use the iPhone image as input to swap in the new product it takes some initial setup but once you have the workflow dialed in for that jewelry brand it becomes repeatable for every new product are you comfortable with ComfyUI or would a simpler hosted tool work better for your situation

Martin Mufaya

Apr 18 •

🛠️ dev-help

Help please, am stuck

Hello. I'm connecting a text parser to slack in order to obtain just the important text from the message in the photo, but it seems i can't get the right formater to put in the pattern field. Please help

New comment Apr 24

Asifuzzaman Arko

0 likes • Apr 18

@Martin Mufaya glad you got some help on the JSON parsing side one more thing worth knowing for future workflows like this: if the message format is inconsistent, meaning different people write it slightly differently each time, regex and JSON parsing will both break eventually a more resilient approach is to pass the raw message to an AI node and just ask it to extract the specific fields you need in a structured format, it handles messy natural language much better and you don't have to maintain a pattern every time the input changes have you got the workflow fully working now

1-10 of 44

Level 3 - Builder 🛠️

42points to level up

Asifuzzaman Arko

@asifuzzaman-arko-5354

Driving Business Growth with AI Solutions | Automation Expert | AI Enthusiast

Active 1d ago

Joined Oct 27, 2024

Dhaka, Bangladesh