Activity
Mon
Wed
Fri
Sun
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
What is this?
Less
More
Clief Notes
56.3k
5.0
Free
15 contributions to Clief Notes
Lyceum Learnings 2
Following on from my other question post. I sort of understand the who, inputs, outputs and outcome concept. My next knowledge road block concept is why build in the first place? Some of the challenge questions were maybe someone has already built it so why re-invent the wheel. I am trying to somehow get AI to have ready made challenge questions from both the technical, cost of not doing anything to business pain points a build will address. But even with an ICM, AI seems to have a limited way of challenging a topic. I always have to prompt it to consider other views and insights. It is not yet ready to be full devil's advocate that can apply an expansive counter arguments to a build topic. This is what I am building, (at the moment just to help me personally) - short summary by AI ============================================================================= Telnyx Voice AI + CRM-Agnostic Engine (Twenty CRM & Salesforce): The Mid-Market Revenue Copilot Most sales tools fail because they focus on the wrong problem: spamming cold dials. In the small-to-mid market (10–250 employees), CRMs become expensive, static graveyards of neglected data. The true hemorrhage across a 5- to 20-rep team is deal slippage, unpaid consulting, post-demo ghosting, and 35%+ of sales payroll wasted on manual data entry. This build does not eliminate the CRM—it dramatically enhances and activates it. Connecting Telnyx Voice AI directly to your existing CRM (Salesforce Enterprise or Twenty CRM) turns a passive database into an active, real-time execution engine, eliminating the need for fragmented, expensive sales engagement add-ons (Outreach, Gong, Orum): The Multi-User Mid-Market Economics (Why It’s an Expensive Problem): • The Team Payroll Leak: Across a modest 5-rep B2B sales team ($80k loaded salary/rep), reps lose 2.5 hours/day to data entry. That is 12.5 hours/day—the equivalent of burning 1.5 full-time salaries (~$120,000 AUD/year) on non-revenue admin. For a 15-rep team, that payroll waste exceeds $350,000 AUD/year.
1 like • 3d
@David Vogel @Aurel Babiš - Thanks for the suggestions. I will start practicing to build my memory muscle. I was looking too hard outside but I just needed to be a bit more critical of the AI answers to challenge myself and the AI response.
0 likes • 10h
@David Vogel @Aurel Babiš - it looks like I will have to build a Shark Tank ICM. This will be interesting 😃
Lyceum Learnings
Hi All, Just thinking out loud as I am accessing the Lyceum recordings instead of attending the live event but not sure if skool is the next best thing since the recordings does not have Discord server access. One of the foundation knowledge Jake and his team is making sure an ICM user understands is knowing and qualifying what to build before touching any AI tool. Fair enough - not many build houses without at least planning the details. The process is, in my own words, 1. Know your target audience/user - what are you trying to build for this group, are you trying to solve a problem, help speed up their process (is this really bothering them) - book keeper entering spreadsheet numbers to an accounting system 2. What materials or inputs do you have to help solve for the user - status reports, where do these users normally start when they do the work - the spreadsheet of numbers 3. You have to convert the starting materials to a useful outcome - numbers entered into an accounting system 4. If I deliver 2 and 3, then accounting reports can be seen faster by decision makers and business owners. Is the chain of thought correct. Can the community who saw the Lyceum suggest other examples?
1 like • 4d
@Charles Dove I tried to upload to Gemini Notebook using the Google Drive link but no luck. Can you please list the steps? I could be missing a step.
3 likes • 3d
The were a lot of additional learnings I got out of the onboarding session. Shout out and thanks to @Nati Peek @Jeff Van Leenen @Brody Billings for your insights. Here is the meeting capture as it is a lot of info packed in one session. It is an AI transcription so it cannot spell everything correctly. https://au.meetinginsights.com/mi/external-player?confid=6afe39a0ec516aaf3c2afe60&subject=Clief%20Notes%20-%20Onboarding%20Meeting&tenantId=2fd52312-8761-4781-a2ea-1217c9e41a6b
I sent my AI for lunch... It came back sure of itself, and wrong.
"If I asked you to go get my favorite lunch, then left the room, what would you do?" I asked that yesterday while teaching a small group of staff who aren't techies. You'd guess. You'd come back with something, and you'd be sure of it. That's what AI does when you don't give it the details. So I turned it into a 90 second Bip short. Bip (8 bit, Zelda inspiration...) is my robot that stands in for the AI. It's 3 steps: give the details, say what done looks like, (thanks @Ari Evergreen !), check and fix. On his second trip Bip checks the bag against the list before he hands it over, finds the peanuts missing, and goes back. The script went through a loop too. I gave Perplexity (a search AI that shows its sources) @Bas Rosario loop shape: 12 pass or fail checks, 3 tries at most, every try judged against the original list. My version passed 10 of 12. Its first try passed all 12 and added the line I like best: "Your quote, your email, your job list. Same 3 steps." It also wrote a prompt to copy. I haven't run it on a real job yet, so try it and tell me what breaks: Step 1: Give the details. My job: [your job]. Ask me up to 5 questions about anything you need to know first. Step 2: Say what done looks like. Write a short checklist of what done looks like for this job, then wait for me to OK or change it. Step 3: Check and fix. Before you show me anything, check your work against the checklist, line by line. Mark each line pass or fail. Fix every fail, check again, and only show me the work when everything passes. What's the job you'd try it on first?
I sent my AI for lunch... It came back sure of itself, and wrong.
1 like • 3d
@Jeff Van Leenen - Thanks for this. I attended the onboarding and got to see you online in person. Thanks for all your input. It looked like your businesses is in the building industry. I am curious what types of problems you have been solving using ICM. Thanks again for your generous insight. I just hope none of the AI shops decide to buy @Jake Van Clief and team!
Online Community Meetup
I attended the Community Meetup and I just want to quickly ask if people who attended are happy for me to share the transcript link.
0 likes • 5d
I was only able to get capture until I had to drop off and attend another meeting. But the discussion around the different projects people are working on are there.
Onboarding call mix up!
Sorry for everyone who joined the onboarding call today and didn't see anyone. It was supposed to be scheduled for Wednesday not today! I was in the middle of a call with a client and didn't notice the mix up untill ten minutes too late! Link is updated and I will be there and even answer some questions in general about your builds and such as an apology ! Forgive me for my negligence !
3 likes • 5d
@Jake Van Clief All ok. These things happen. Looking forward to the reference links. I was just going to ask if there is a separate Discord channel for Edubaware.
1-10 of 15
Joseph Dizon
3
5 points to level up
@joseph-dizon-8945
Joseph, Sydney. Head of IT at an MSP — deploying AI voice, Copilot, and contact-centre AI for SMB customers. Here to learn and build.

Active 10h ago
Joined May 18, 2026
sydney, Australia
Powered by