Used astra for a voice agent in todays build and thought id share. It broke three ways before it worked: 1. It said "give me a moment" and stopped. 2. Told it to keep talking. It started guessing the answer. 3. Tightened the prompt. It faked "here is what came back" with the wrong number, 20 seconds early. Fix: two nodes. The desk node calls the research tool, and that call moves the conversation to a waiting node whose only job is one sentence and stop. It never saw the question, so it cannot guess. When the real answer lands, the waiting node delivers it and hands back. The thing I keep: if a rule has to always hold, put it in the structure. A prompt is a request. A node boundary is a wall.

Quick before/after on a project I've been heads-down on: Before: A lodge owner manually answering every booking call, hand-calculating quotes, texting door codes to guests one by one, no idea what dates were actually free until he checked three different places. Now: A guest picks dates on the site → sees the exact price, computed server-side and locked so it can't be tampered with → submits a structured inquiry that lands straight in the CRM → owner approves with one tap → Stripe invoice goes out automatically → guest gets a unique, auto-expiring door code the moment payment clears → after checkout, they get a review link → their feedback sits in a queue until the owner approves it → it shows up on the live site. Eight separate systems, all handing off to each other, all built on real database logic instead of hardcoded values anywhere. None of it required the owner to learn to code. It required someone (me) willing to sit in a terminal at 2am reading Deno error messages until a function stopped returning 503s. The unglamorous truth about "AI-built" systems: the AI writes fast. The person still has to know what to ask for, why a security rule matters before something breaks, and when "it works" isn't the same as "it's actually safe to hand to a paying business." That gap is the whole job now. Everything else is just typing.

I’ve built a social media automation skill into my AI-OS, 2 months ago and I’ve left it run untouched for the past 6 weeks and now I’m starting to feel the power of it. What it does? I give it my skool where he constantly scraps my videos, edits them with a badge and CTA outro, puts them in a folder of approved videos that has access 24/7 in the cloud. It checks the files for Visual QC and post-worthiness, the rejected ones get put into a folder where it need my attention. then publishes the right version to Instagram, TikTok, Facebook and YouTube with captions made for each platform separately and accordingly to the content of the video so each video gets a different bespoke caption. After that it checks the posts are actually live, saves the links and stats, then moves the video out of the queue and into a new folder so it doesn’t get used twice. It does this 3 times a day for 4 platforms (but I can add more platforms if I need to) It handles scheduling, automated 1st comment on each post, comments, keyword replies (like manychat) and sends DMs with button triggers and links, it also does account health-checks, post analytics & follower tracking. Learns from it and can give suggestions on what worked better visually or what hours are best to post..which comment needs my attention or just replies automatically if I want to. Since switching it on: It posted 392 posts 81.9K views 1.1K interactions +53 followers It’s gathering a huge footprint for my business across multiple platforms and only growing Ps: The accounts where all brand new so 0 audience while Facebook and YouTube having only been introduced 2 weeks ago. So they have been running less posts compared to Instagram and tiktok. Essentially this gets me 360 videos a month posted and today i've introduced the 5th platform [Threads] Which will take this to 450 posts a month - 90 original videos - 15 posts a day. Still improving bits as I go, but it’s already saved me a unfathomable amount of repetitive work and it’s compounding as we speak.

I like projects that start with: “We know what we want to build, but we’re not sure how to build it properly.” I’m a senior software engineer and most of my work sits somewhere between full-stack development, AI, automation, mobile, and backend architecture. A few examples of things I can build: → An AI agent that reads documents, makes decisions, calls APIs, and takes actions instead of just answering questions → A SaaS product from rough idea to a working MVP with authentication, subscriptions, admin tools, and deployment → WhatsApp/Discord/web assistants connected to CRMs, internal data, calendars, and business workflows → Existing systems that need to be untangled, stabilized, sped up, or prepared for real production traffic → Mobile + web products sharing the same backend → IoT/robotics systems connecting real hardware, sensors, cloud services, and AI My usual stack includes TypeScript, React/Next.js, Node.js, Python/FastAPI, PostgreSQL, Redis, React Native/Flutter, AWS, Docker, and LLM/RAG tooling. I’m not only looking for neatly written project specs. If you have an unfinished prototype, a half-working product, or even an idea where the technical direction is still unclear, send it over. I’m happy to look at the problem first and tell you how I’d approach it. No sales call required.

If you got a brand new lead, and they wanted a demo voice agent for their business, what stack would you use and how long do you think it'd take you to set up? Say receptionist, for example