Launched this week
Tadka
Ship 10x more ad creative, without hiring a design team.
102 followers
Ship 10x more ad creative, without hiring a design team.
102 followers
Tadka turns one brief into hundreds of on-brand, audience-tuned ad creatives in minutes, then learns which ones win. The creative volume your Meta and Google campaigns are starving for.
Core Features
Brief-to-Creative Pipeline — input brand details, products, target audiences → generates on-brand ad variations
Multi-channel Export — Meta, Google Ads, TikTok, Pinterest, email, landing pages (PNG/WebP)
Visual Styling — Bold & Vibrant, Muted & Editorial, Soft & Minimal, Warm & Playful
Performance Learning Loop — tracks ad performance, auto-shifts toward converting styles
Product Integration — syncs with Shopify and WooCommerce
I attempted to run the process expecting it to generate eight different variations of my image, but only one was created. I also encountered some bugs during generation. For instance, while I can upload multiple images, the server returns an error for more than 4 images I think. Additionally, I can add images but cannot remove them, after accidentally adding the wrong image, I was forced to restart the entire process to upload the correct ones. Also, review the payment modal the padding is missing.
@matheusdsantosr_dev really appreciate this Matheus, this is exactly the kind of detail that helps. all four are fixed now: the variation count (you should get all 8), the multi-image limit, the inability to remove an uploaded image, and the payment modal padding. so you can swap out a wrong image without restarting the whole flow now. would love for you to give it another run, especially the generation and checkout, and tell me if it holds up. and thanks for catching the payment one, that was the important one.
@nawneet77 happy to help! Just gave it another run and everything looks solid now. Congrats on the launch!
The brief to creative gap is where most teams lose time. Does Tadka pull from competitor ads at all or purely original
generation?
@mohamed_hussein25 Good eye, Mohamed. It's mostly original generation from your brand inputs, but we do factor in category/competitor signal to understand what's working in your space used as direction, not for copying creative. The goal is on-brand variety, then the performance loop sorts winners from there. What's your current process, do you study competitor ads before briefing, or start from scratch?
Receiptor AI
Congrats! What kind of creative can it creates? static visuals? videos?
@luigi_receiptorai Thanks Luigi! Right now Tadka focuses on static visuals ad creatives exported for Meta, Google, TikTok, Pinterest, plus email and landing page formats (PNG/WebP). The whole idea is generating a high volume of on-brand static variations fast, then learning which ones convert.
Video's the natural next frontier and it's on the roadmap, but we wanted to nail static volume first since that's where most teams are starved for creative.
Ad creative is one of those jobs where speed matters more than everyone admits. How do you keep outputs from looking like the same AI ad soup?
@sarveshsea this is the thing I cared about most, honestly. the soup problem comes from everyone prompting the same models with no constraints. Tadka starts from your actual brand inputs (your products, palette, voice, audience), so generation is bounded by your identity instead of generic. then the performance loop kills the variants that don't land, so over time you're not just generating variety, you're converging on what works for your specific audience. variety with a filter, not variety for its own sake. how are you thinking about it on your end?
the 10x without a design team framing is what every growth team wishes for. the hard part is usually brand consistency once volume goes up. curious how Tadka holds the brand together across 100 variants. is there a human in the loop reviewing tone and color before they ship, or do you encode the brand once at setup and let the model self police inside those bounds?
@thenameisarian great question, this is the exact tradeoff. the way we hold it together...you encode the brand once at setup (palette, type, voice, dos and don'ts), and generation is constrained to stay inside those bounds rather than free-styling each time. so the 100 variants are variations within guardrails, not 100 independent guesses. there's an optional human review before export if you want the final say, but the point is you shouldn't need to eyeball all 100. curious from your side: at what variant count does brand drift usually start biting for your team?
Hey everyone, Nawneet here, maker of Tadka.
Real reason I built this is because I kept watching good campaigns stall. Not because the targeting was off, but because we could only ship 3 or 4 creatives a week. Meta and Google want volume, and a small team can't feed them fast enough. Hiring designers for that is slow and expensive.
So Tadka takes one brief and turns it into hundreds of on-brand variations in minutes. Then it tracks which ones convert and leans into those styles.
Would love your take. What's your biggest creative bottleneck right now: volume, testing speed, or staying on-brand?