... than with my own wife We went from two strangers in an Antler cohort to running a profitable, bootstrapped company. Pivoted a couple of times (gets less painful as you go), built an enterprise B2B feedback repository platform @Dolphin AI that got product 5 of the day, a B2C/Prosumer AI tool @Pretty Prompt that got product 2 of the day TWICE , scaled to thousands of users, and billions of AI requests purely PLG. We kept it fully bootstrapped with ...

... your biggest pain points when web scraping or managing social media campaigns at scale? What tools or anti-detect browsers do you integrate your proxies with? (We currently integrate out-of-the-box with tools like Undetectable, Octo Browser, Dolphin {anty}, and AdsPower). Looking forward to your feedback, suggestions, and feature requests! #gonzoproxy ...

... Friday I took my first salary as a founder. Paid by revenue. Not investment. My first one ever. So proud. Most startups die before they make $1. We ve been making revenue for 8 months. Even after pivoting from @Dolphin AI . I don t take this lightly. But it isn t about money. It s about the milestone. The fact that something didn t exist and now it does. Something people want. Are happy to pay for. And happy ... ... Glukharev(@alexeyglukharev): @ilaiszp  my appreciation! You re great! I would know the full detailed story:) Comment from Swati Paliwal(@swati_paliwal): Congrats on that first revenue-funded salary; super proud of you. Huge milestone, especially post-pivot from Dolphin

... auto-detects your CPU/RAM and recommends the best fit. Supported GGUF models with Q4_K_M quantization include: Qwen 3.5 (0.8B up to 9B) - Llama 3.2 (1B, 3B) - Qwen 2.5 Coder (great for writing code locally) - Dolphin (uncensored variant) One-tap downloads from HuggingFace with background management and RAM warnings before large downloads. The Chat Experience Real-time token streaming (words appear one-by-one as the model generates) - Conversation history saved locally - Image input support (multimodal) - Local Speech ... ... 31MB size is a huge advantage . Most people assume local LLM = heavy and slow. You re breaking that assumption. The phone automation agent is probably your most underrated feature. If it becomes reliable, that s a real differentiator vs

... visit my family for the New Year, I scrolled through my phone and realized how much happened since we launched Pretty Prompt on Product Hunt 7 months ago. Here s a little photo-dump recap : 1 Pivoted from @Dolphin AI to @Pretty Prompt (super scary at first ) 2 Grew from 0 19,000 users (and counting!) 3 Built something people genuinely love ( 200+ 5-star reviews ) 4 Shipped 166 product tickets (and many more bugs ) 5 Took a trip ...

... perfect, but it's real. Raw, honest, and down to earth. Not some shiny I m a founder, and everything is great type of thing that you all the time... It is quite the opposite: - How we pivoted from @Dolphin AI @Pretty Prompt - Why we decided to burn the boats and go all into this - How we see Pretty Prompt as a Context Layer for AI - The Launch on Product Hunt ( 2nd product of the day ) - Users asking ...

... iPad I feel like it s building skills: Creativity & narrative: combining multiple stickers into stories Language development: I ve made her use only English so she s learning how to make clever phrases and sentences like jumping dolphin over a wide river Iteration: tweaking prompts (what she says) when results don t match intent Fine motor skills: coloring and sticker placement Patience and focus: focus on her idea, be patient for the output, and focus again on the coloring ... ... teach her some to create an image for Stickerbox makes me really happy to see. It s a rare tech gadget that earns a place in the quiet-time toolkit and actually helps my kid focus. Comparisons & Alternatives

... library that keeps a language model in training mode while it chats. ARC Core uses LoRA adapters plus elastic-weight-consolidation to update weights on the fly, then sleeps every N turns to consolidate memories. Key numbers on Tiny-Dolphin-8B (30 evaluation rounds): 37 % perplexity (fluency) 54 % median latency +8 % coherence score It is one pip install metisos-arc-core Github Repo: https://github.com/metisos/arc_c... Apache 2.0 License license, no server dependence. Full benchmarks, code in the README. Feedback ...

... something else? P.S: I'm not in a dire need of such things, but I won't mind hearing about some that you guys have used before. ("-") Comment from Alex Lou(@fullstack_x): Using AI to talk to dolphins. https://blog.google/technology/a... Comment from Siddharth Pereira(@siddharthpereira): There s a guy in India using to AI to analyze voice patterns to understand how his infant baby communicates with his dog. Completely private usage product that google noticed and is eyeing ...

... mostly. Sure, a lot depends on many things, but i didn't have any good experience with a game dev yet for me, "money in game dev" seems like the "Survivorship bias" We only know stories about saved by dolphins only, because dolphin