Abhinav

@ashishllm @vasuman Oh understood! I did Integrate Coqui TTS and it is indeed fast! But I did read on reddit that they've shut down, so future support might be uncertain. Anyways, thanks for the suggestion!

@shipitabhi @vasuman I also used RealTimeTTS library and used coqui engine under it

Got over 150 DMs from people who want to work for my company as an Applied AI Engineer. Cant possibly interview you all, so I’m going to do it this way: - comment with your best work. I’m not looking for a flashy frontend, I want to understand what workflow you solved for (can be complex agents, or simple image generation) - describe how you prompted, what worked for you / what didn’t, and what edge you have over the other 10000 “AI Engineers” who call an API and call it a day I’m going to be very clear about what I’m looking for: I want people who test every single model, and learn through first-hand experience which models serve which use cases, beyond just “Claude for copywriting, Gemini for image generation”. Ideally you can productionize the workflow through code, but even that is a nice to have. Specifically looking for image and video generation experts. Everything else is lower priority but still needed. For the top 5-10, I’ll open up a mentorship cohort, where I can help you with whatever you need, share some advice on what to build and how to succeed in this space (completely free), and more so just observe how you work and what you build over the span of a week. Top 1-2 gets the job. It’s part time and contract based, paid hourly. Should be 5-10 hours a week minimum, more if needed. If that sounds fair, again, comment below what I asked for above. This isn’t for everybody and that’s fine, I want people who align with the vision. Location does not matter. Thank you.

@ashishllm @vasuman Will check it out, thanks!

@shipitabhi @vasuman try Coqui TTS with GPU once and do let me know

@ashishllm @vasuman Hey @ashishllm Unrelated to the main thread, but What Open-source TTS model did you use? Building a very similar project here Have tried using:- Orpheus TTS: High latency Dia TTS and Chatterbox (Better than Orpheus, but high latency) Piper: Low latency but robotic voice.

@vasuman 2) Built ultra low latency Ai Voice Bot for Healthcare niche, on top of AWS, with complete end to end open source models, for TTS and LLM generation, and for STT, used AWS transcribe. TensorRT optimization is being done to achieve 40 ms of First word token. Blazing fast 🔥🔥

@p_roguszewski Yup planning to do that!

It's Friday day.... Are you cooking over the weekend ❓

Update: Haven't received any reply. Will be following up! Haven't been able to build anything this week. Busy with work.

@TerekhinIvan Good luck!

Starting tomorrow, my GameDev journey officially kicks off. I’ve decided to focus entirely on this one game. Shoot Them Up style shooter with little RPG elements. Vampire Survivors in 3d, I don't know. Tomorrow I’ll do a full review of the project, host it properly (LFS Git maybe), go through the backlog, and get things rolling. I’ve got about 2 months (this summer) to give GameDev another real shot to try and get into the rhythm again, or I will accept my fate as JSON master and play that card. Let’s go! In parallel I have podcasts, and I need to play marketing games for the current apps. I still think that I can do something (in money terms) with Sober Tracker and WinGym. And they are already there. Will keep you updated here!

@coopmacg Until you have bills to pay

@Irem__Victor Yeah often feels like that. I got my first reply after sending out 100+ emails(that was for a job) Maybe you never stop until you get a reply :-⁠)

@shipitabhi Ohhh, okay How was the cold email experience like? It's what I'm doing rn, but it feels I do a lot of work without getting results

Sent a few cold emails today with some website redesign demos I’ve been working on. First time doing this, a bit nervous, but it feels good to put something out there. Will share the links if any of it moves forward

@byLuocca I got more clarity by the 2nd one! :)

@shipitabhi Really? How come you like the second one better?

same architecture, different diagram Which one do you prefer?

@dhmn__punit That's great! keep going!!

one of our clients had 0 users for 2 months we stripped their app to one feature and slapped on a share button 8 users in 24 hrs 1 paid MVPs don’t need more code they need vision

@lushythedev Yeah that sounds cool!

@shipitabhi I was thinking of something like this. Except replacing the progress ring with shapes, that get longer and more complicated the higher your streak is. Thoughts?

@RelaxKartikey Yup, following you!

@RelaxKartikey Oh that's awesome man! Just focus on the message/email you send Make sure to draw their attention(imagine them filtering out 100s of DMs) The age of the account won't matter in this case. Good luck bro!

@RelaxKartikey Firstly, Welcome to X! It's great that you've prior experience with clients, that'll give you a great head start. You can try to build a portfolio(if you don't have one), and try to Cold email/cold dm founders. post about your work on X. AI is your friend here 😉

@mhdzaaaid @boltdotnew @AnthropicAI +1 Ran out within 2 hours🥲

@boltdotnew @AnthropicAI What i ran out of credits

🥁 Introducing the One-Shot Challenge Bolt is FREE this weekend for EVERYONE, thanks to @AnthropicAI! → Create a stunning app in a single prompt → 25+ winners will take home $1k each → Over $1M+ in other hackathon prizes, jump in now! Lezzgooo!! Here’s how to get started:👇

@llrnsss Nah you gotta yap at it😆

So you are telling me that if I ask things politely to Cursor he won't work better??

@Irem__Victor Had done cold emails previously, but it was for a different purpose

@shipitabhi First time doing cold emails? Just curious

Sharing few screenshots below👇