What changes when a retailer can make a separate decision for every customer instead of placing thousands of people into the same campaign segment?

In this episode of Startup Builders & Backers, I speak with Ilyas Kurklu, CEO and co-founder of Replenit, about the company’s attempt to move retail CRM beyond fixed audiences, campaign calendars and generic promotions. Replenit calls its product Maestro, an AI CRM Manager designed to decide what should happen next for each customer and pass that decision to the retailer’s existing channel tools.

The distinction between prediction and decision runs through our conversation. A propensity score can estimate what someone might do, but a retailer still has to decide whether to act, which product to recommend, when to communicate and whether a discount is commercially sensible. Ilyas argues that this reasoning should be readable by a human CRM manager. For every decision, the team should be able to inspect why this customer was selected, why the timing was chosen and why a particular product was recommended.

That raises the question every enterprise AI project eventually faces: how do you prove the system created value rather than claiming credit for a purchase that would have happened anyway? Ilyas recommends measuring incremental margin per customer against a randomized holdout. He is skeptical of attributed revenue, influenced revenue and open rates when they do not reveal whether the business made additional money. We also discuss Replenit’s focus on cost per decision and the processing expense that comes with adding further data to an AI model.

Integration is another part of the story. Retailers may already use a customer data platform, a customer experience platform, analytics products and multiple communication channels. Replacing that stack can bring cost, delay, security reviews and regulatory work. Ilyas explains why Replenit has been designed as a reasoning and orchestration layer that works with the systems a retailer already owns.

The conversation also considers customer interest and commercial pressure. A retailer may have contractual brand rules, margin targets and limits on which products can be recommended together. Maestro records brand, product and customer memory, while the retailer sets the rules that govern its output. The human CRM manager then becomes responsible for reviewing outcomes, setting boundaries and questioning the system’s decisions.

For founders and backers, Ilyas offers a refreshingly practical reminder. Building an AI company still means doing the ordinary work of running a business, proving that the buyer can make or save money, answering why an incumbent cannot easily copy the idea and reducing the risk that stops an enterprise customer from saying yes.

Would you trust an AI CRM manager to make individual customer decisions, and what evidence or controls would you require before giving it that responsibility? Share your thoughts with me at Tech Talks Network.

What can founders learn from a healthcare workflow that large organizations and independent clinics were both struggling to fix?

In this episode of Startup Builders & Backers, I speak with Charlie Jordan, founder and CEO of VITL, about identifying overlooked customer problems, building technology for legacy markets, and making complicated systems feel simple without weakening trust or compliance.

Charlie’s route into healthcare began during the rapid growth of telehealth in 2020. He later saw compounding pharmacies face a sharp rise in demand, driven in part by direct-to-consumer healthcare and GLP-1 treatments. Fulfillment systems that worked at one volume could quickly develop long backlogs. Prescribers were also entering the same information repeatedly, sending prescriptions into workflows with limited visibility, and spending time chasing pharmacies for updates.

The business opportunity became clearer when Charlie moved from large telehealth organizations to independent practices. The smaller clinics were experiencing the same problems but had far fewer resources available to solve them. If organizations with large budgets and practices with small budgets were both stuck, he reasoned that the workflow itself needed a different answer.

VITL connects cash-pay clinics with verified compounding pharmacies and gives practices a single place to prescribe, compare options, and follow fulfillment. Charlie says the platform has reduced the initial prescribing process from roughly four and a half minutes to 30 seconds. He argues that the larger saving comes after the prescription is written, when status checks, problem resolution, shipment updates, and patient communication would otherwise create hours of additional work.

Speed, however, cannot be separated from responsibility. Compounded medications may carry serious consequences, and Charlie believes the market needs clearer ways to describe and verify quality. He explains how VITL reviews pharmacy standards and ingredients before presenting practices with a network they can use. The aim is to make the experience easier while keeping the checks that protect clinics and patients.

We also discuss Charlie’s design philosophy of beautiful simplicity. His teams ask how little information is needed to complete a task and how few steps a workflow can contain. Repeated data entry is a warning sign. Information collected during patient intake should move through the prescribing process rather than being typed again across several screens.

For founders, Charlie’s most direct lesson concerns uncertainty. Failure is part of building, and no founder will know the answer to every question. The skill is choosing a direction quickly, gathering evidence, and changing course when the decision proves wrong. That lesson becomes especially important in regulated markets, where progress must balance speed with responsibility.

The supplied company brief says VITL works with over 700 clinics and announced a $7.5 million Series A led by SignalFire in March 2026. The scale gives Charlie’s discussion a practical context, but the wider idea applies to any founder facing a market weighed down by old systems and manual work.

Where has your industry accepted unnecessary complexity, and what would it take to redesign that experience around the people doing the work? Listen to the conversation and share your thoughts with me.

What happens when founders can build faster than business buyers can adopt?

In this episode of Startup Builders & Backers, I speak with YouMon Tsang, founder and CEO of ChurnZero. YouMon has built four companies and completed three successful exits, giving him a candid perspective on category creation, B2B market timing, AI adoption, and the confidence required to continue when the outcome remains uncertain.

YouMon believes entrepreneurial founders often run ahead of the market. They see the technical opportunity, create a new category, and then discover that business customers need time to understand the problem, assign an internal owner, approve a budget, and decide what evidence they require. He now sees a strong case for arriving second or third, when buyers recognize the category but remain open to a better answer. It is a useful counterpoint to the startup belief that being first automatically produces a lasting advantage.

We also discuss how AI is changing customer success. ChurnZero works with customer teams focused on retention, expansion, and customer experience. Automation has supported go-to-market teams for years, but agents can now handle decisions and tasks that once required manual review. The important question is when software should recommend an action, when it should complete that action, and what information it needs before the result can be trusted.

For YouMon, the answer begins with data. He separates customer data into facts and information. Facts include contract dates, spending, and points of contact. Information includes call transcripts, emails, product documentation, and knowledge about how the customer uses a service. An AI system needs both to produce a useful recommendation. A clean account record cannot compensate for missing conversation history, while a detailed transcript is of limited value if the system has the wrong contract or contact data.

That combination also affects timing and empathy. Customers quickly notice an automated message that arrives at the wrong moment. YouMon argues that poor timing usually exposes missing information or a badly designed workflow. The language of an email can be tailored, but an empathetic message sent at an inappropriate time remains a poor customer experience.

Trust must be earned gradually. YouMon compares today’s AI agents with the earlier adoption of rule-based automation. Teams initially asked a person to approve each message, then removed that checkpoint after the workflow repeatedly behaved as expected. Organizations can take the same approach with agents. Some may permit immediate action, while others will keep a person in the loop until the business has enough evidence to feel comfortable. The right boundary will vary by company, customer, and consequence.

Founders selling AI must also recognize that buyers require different forms of proof. An early adopter may try an interesting technical answer with limited evidence. A fast follower may want a strong demonstration and one or two customer examples. The wider market will expect social proof, controls, and clear evidence of success. A founder therefore needs to understand where the buyer sits before deciding which argument, metric, or reference will earn confidence.

The conversation closes with YouMon’s view of entrepreneurial uncertainty and a personal story from a difficult fundraising period. Before another investor meeting, he looked in the mirror and changed the question from whether someone would invest to which investor would be lucky enough to join the company. The business had to earn support, but his posture changed from asking for permission to presenting an opportunity he believed had genuine value.

Do founders gain the greater advantage by creating a category, or by arriving later with a better answer when customers are ready to buy? Listen to the episode and share your thoughts with me.