Most people don't like doing testing, so we're trying to help. We're doing a major upgrade to our voice ai automated testing platform and we're looking for feedback on: - how you do testing now - what tools you use (if any) - what do you like about them - what do you NOT like about them - what do you wish they could do better - any other pain points in your development process Since this is a voice ai group, and nobody likes typing long answers, we've made a cool Voice AI Testing Interviewer that you can just talk to (or chat with if you don't feel like talking out loud) at https://interview.lemalogic.com/testzilla-discovery . We will take all your suggestions and pain points into consideration for the next release and our future roadmap. You'll also get an emailed transcript of your interview for your own records. Thanks for letting us know how to make things better for all of us! Cheers, Brian

Our Isle Of Man AI meetup group is having a Christmas AI Contest… so we built SantaCalls.ai, a free “Talk to Santa” app for the island's kids. If you’ve got kids (or are still one) try it and let me know what you think. We've tried to put pretty strong guardrails on it, as it's going to be kids interacting with it, and designed it for safety first. But if you find anything we missed, please let me know (and bonus points if you have a way to prevent the problem from happening). Entirely vibe-coded with Claude Code, I didn't write one single line of code, though I did monitor and "curate" its work as it was building. But it did 95%+ of what I asked on it's own. We built the first prototype in less than a day, built out additional features over the next 2 days, and have it running completely in CloudFlare for web hosting and Retell for Voice and OpenAI for, well, AI. We've created a master agent called "Boardroom" that I've been training to do all kinds of things. It was able to go create the website, the brand guides, the Retell prompt, automatically create the Retell Agent and links it to phone numbers (we haven't released the phone version yet, but it works), create all the images and font choices, setup all the hosting and workers on Cloudflare (nothing is on our own servers), configured the DNS entries, manage the repos, setup github actions to deploy updates to Cloudflare, connected it to Resend (for email sending), wrote all the backend code, our DPIA, privacy policy, searched the internet to build the Gift Ideas section, built the search interface, and, well... pretty much everything else... all from telling it what to do in Boardroom. It deployed agents and workers to do everything asked. It also maintains all our documentation, handles our Asana tickets, and writes daily standup reports, amongst other things. Totally amazing new way of working, at unprecedented speed of deliver. I also built four other websites at the same time, in four different git repos, and Boardroom managed everything pretty flawlessly.

Cal.com has launched their own voice ai scheduling agent, Cal.ai You can see it on Product Hunt at https://www.producthunt.com/products/cal How do you see these tools that used to be integrations (like 11Labs) now becoming competitors in the Voice AI space?

By Brian Gallagher, CEO of LEMA Logic & Testzilla.ai Introduction This guide is a combination of information from my previous posts in the AAA group, comments and responses to questions, and a recent webinar I gave on Voice AI pricing strategies (slides attached). As the voice AI space continues to evolve, one of the most common questions I get is: "How should I price my voice AI solutions?" Here, I've compiled insights on pricing strategies that have worked well for both our consulting services and SaaS products. I'm Brian Gallagher, owner of LEMA Logic, an agency with decades of consulting experience ranging from startups to Fortune 500 companies, government agencies, and educational institutions. We’ve been around long before GenAI was on the scene, and we’re having a blast creating new solutions using AI to do things that weren’t possible just a few years ago. I'm also the developer of Testzilla.ai, a Voice AI Automated Testing platform that provides easy quality-assurance (QA) testing of voice AI applications. Our platform uses AI to test your AI, eliminating the tedium of manually testing voice applications while ensuring consistent quality. Pricing isn't just about making money—it's about positioning yourself in the market, communicating value, and building sustainable business relationships. Let's dive into the strategies that can help you price your voice AI solutions effectively. Why Pricing Matters (Beyond Just Revenue) While pricing obviously matters for revenue, it's important to understand the other impacts it has: 1. Market Positioning: Your price point immediately signals where you sit in the market. Apple has positioned itself as a premium brand, allowing them to charge significantly more than competitors for similar functionality. 2. Perceived Value: Pricing affects how clients perceive the value of your offering. Lower-priced solutions are often perceived as having less value, even when technically equivalent.

[I originally wrote this as a reply to a post, but since many people expressed interest in the topic, I’m making it a standalone post to help others with similar questions.] A common question is, "How do I price my Voice AI services?" There are a few different strategies you can use for this. The most common one is Cost-Plus Pricing: Cost-Plus Pricing: Here’s the basic math: - Cost per minute * 60 = cost per hour - Example: VAPI @ $0.16/minute * 60 = $9.60/hour cost If we’re selling a 10-hour/month plan, our cost is: - COST: $9.60/hour * 10 hours = $96.00 per 10 hours per month Then we add our markup. A typical human call center can charge $1.00/minute or more, so the cost of 10 hours of a human call center service would be: - HUMAN COST: $1/min * 600 minutes (10 hours) = $600 per month for a human call center. So, we could apply a 400% markup and still be cheaper. A 400% markup means the selling price is 4 times the cost. For every $0.16 in cost, the selling price would be $0.96, resulting in an $0.80 profit. ($0.96 price - $0.16 cost = $0.80 profit). - OUR PRICE: $96/mo cost * 400% markup = $384 per month Customer Price. Now let’s add some pricing psychology. Let’s round it up to $400, then use the "ends with a 7" trick to make it $397/mo, so it "looks" less than $400. So, we have a monthly price of $397, which costs us $96, resulting in a profit of $301 per month per client. PACKAGING YOUR PRICING: Now that you have your cost and price, let’s package it up. You could structure this in a few different ways for a customer contract: 1) Monthly Fee with Maximum-Minutes Cap Offer something like: "Our Starter service costs $397/month and includes up to 600 minutes (10 hours) of inbound calls per month. Any minutes above 600 are billed in 60-minute (1-hour) blocks at $60/hr." Benefit: If the customer uses more than 10 hours, you’re covered and make more money. But if they use under 10 hours, everything they don’t use is pure profit for you.