And how founders can build one
Over the past quarter, nearly every founder I meet has been exploring ways to bring AI into their workflows — not as a flashy feature, but as real leverage. The CRM is usually the first place they look, and for good reason: it’s where data, follow-ups, and accountability lives.
Every software company runs on customer data, but most still manage it through a patchwork of manual workflows — logging calls, updating notes, chasing follow-ups, and pulling reports. These are necessary but repetitive tasks that drains time, teams, and lacks consistency.
A CRM agent automates those tasks. It’s not a chatbot; it’s a background system that connects to your CRM, reasons through what needs to happen, and takes action. The result is less admin, faster cycles, and cleaner data. In this post, I’ll break down what a CRM agent actually does and share a simple roadmap any founder can use to build one — no AI research team required.
A Simple Example
Imagine a rep says:
“Follow up with all leads from last week and personalize the message.”
Here’s what happens behind the scenes:
- The agent queries the CRM for leads from the past week.
- It prompts an LLM to draft personalized emails based on each profile.
- It sends those messages through the CRM’s native system.
- It logs each interaction automatically.
The LLM handles reasoning and content.
The workflow layer ensures ordering and reliability.
APIs provide access to the CRM and communication tools.
Cloudflare outlined this exact stack in its AI presentation earlier this year (see below). It’s the clearest blueprint yet for how agentic systems will interact with enterprise software.
We’ve seen early versions of this inside portfolio companies — one founder built a lightweight agent that automatically logs post-demo notes into HubSpot. It saved reps 5+ hours a week and surfaced missed follow-ups they didn’t even realize were slipping through.
Where the Data Comes From
Frontier models like GPT-4 or Claude 3.5 are generalists. They’re trained on massive public datasets — think the entire indexed internet.
But to make a model useful inside your company, it needs domain-specific knowledge in two areas:
- Workflow data — how your team actually sells, follows up, and reports.
- Customer data — the proprietary context inside your CRM, emails, and notes.
Training a model from scratch is prohibitively expensive, so the practical path is to adapt existing models using three common techniques.
1. Retrieval-Augmented Generation (RAG)
RAG gives your model access to private data — docs, FAQs, CRM records — via a retrieval layer. When a query comes in, the system fetches relevant snippets and injects them into the prompt. Think of it as giving your LLM an open-book test.
2. Fine-Tuning
Fine-tuning uses labeled examples from your domain (e.g., great email threads or sales summaries) to train the model to mirror your tone and reasoning style.
3. Tool Use & Function Calling
The most powerful agents act, not just answer.
Tool use lets an LLM call external systems — CRMs, ERPs, HR databases — to execute real-world actions. That’s how a CRM agent can actually send emails, update records, and sync tasks rather than just suggest them.
How Founders Can Get Started
If you’re a founder, you don’t need a research lab to build this. Start simple:
Step 1: Pick a narrow, high-leverage use case (e.g., “automate weekly lead follow-ups”).
Step 2: Use an off-the-shelf LLM and connect it to your CRM via APIs.
Step 3: Add RAG to ground the model in your company’s data.
Step 4: Layer in human-in-the-loop review so you can trust and improve the system.
Tools like Zapier, LangChain, or Dust make it easy to prototype this without a full engineering team. Start with a single API connection and measure the time saved — that data will justify expanding into more workflows. Once you prove value in one workflow, expand from there. Each agent you build compounds efficiency across the org.
The Bottom Line
CRM agents aren’t futuristic — they’re the next logical evolution of software automation. Founders who start building these systems today won’t just save time — they’ll redefine how their teams operate.
Every company that depends on customer relationships will eventually use one.
The earlier you start experimenting, the faster you’ll compound the benefits: cleaner data, faster execution, and a sales team focused on strategy, not spreadsheets.
Building one isn’t just about AI — it’s about leverage. And if companies like Cloudflare are already running more efficiently because of it, the question is how long others can afford to wait.
A New Era of Tech-Driven Care in Dentistry
One of my first “jobs” was working in my Mom’s dental office some 20+ years ago. The core responsibility I was initially tasked with was writing her ‘recall’ cards – postcards that communicate oral care reminders to patients.
I would write them by hand, and we would batch-mail my output at the end of each workday at the mail annex she co-located with (as an aside, a Mom business lesson, locating adjacent to a post office is an OG growth hack for constant foot traffic and net-new patient growth, but Lessons from Mom should make for a dedicated post for another day)…My starting salary was 10 cents a card! Talk about aligned compensation 🙂
I learned many things from that time in my life, but what remained particularly ingrained in me was just how hard solo practitioners must grind to scale and how painfully manual most providers’ workflows are.
We’re looking to back businesses that change that! The following post was shared on Ansa’s blog on 12/25/2023, touching on some of the angles we’re exploring.
***
As we enter the holiday season, we wanted to highlight an area we have been focused on at Ansa for the past few months: dentistry. Dentistry holds a special place for us at Ansa. Not only is it a field ripe for technological innovation, but it also resonates with us on a personal level. Marco’s mother is a dentist, and Duncan’s grandfather was a dentist. We know visiting the dentist isn’t everyone’s idea of a good time, but that speaks to the opportunity. We believe technology is the key to transforming dental visits from a dreaded chore into a surprisingly pleasant experience.
Ansa’s View of the Dental Tech
Where Dentistry Sits Today
Dentistry represents a significant chunk of healthcare spending in the United States. Taking care of our teeth is one of a limited number of activities we do twice daily (and anyone who follows the recommended 2 minutes twice a day from the CDC will spend nearly 1,000 hours over their life span brushing their teeth), which drives home the importance of dental care in our daily lives.
There are over 185,000 practices and over 200,000 dentists in the United States to service this need. Over 25,000 students graduate from dental school annually. All focused on the importance of teeth and oral health. Dentistry is a huge, growing industry. This is particularly poignant considering the global scale of dental health issues, with an estimated 3.5 billion people facing oral health problems. Unsurprisingly, it is an industry that already supports massive, complex organizations selling into dentist’s offices like Henry Schein (NASDAQ: HSIC) and Patterson. As we explore investing in the space, we have seen a handful of ways to dramatically improve the revenue and profitability of these practices.
Trends Reshaping the Dental Industry
The Emergence of DSOs:
The rise of Dental Service Organizations (DSOs) marks a significant shift in the industry. High dental education costs have made it challenging for new graduates to start independent practices. Over the past 40 years, dentist’s incomes have tripled, but dental education costs have skyrocketed, rising 16 times. This steep increase in tuition makes it increasingly challenging for new graduates to start independent practices, as student loan repayments consume a larger portion of their income.
Many new dentists find joining existing practices, often DSOs, a more viable option. DSOs offer operational efficiencies and access to technological advancements, as highlighted in a 2023 report by the American Dental Association. The industry is witnessing a trend of acquisitions by large, equity-backed DSOs ranging from large consolidators like Heartland and Pacific Dental Services to more localized groups like Singing River Dental Partners. However, despite this consolidation, a significant majority of practices — 93%, according to Harris Williams — remain independent. This balance between DSO-driven consolidation and the persistence of independent practices marks a complex and evolving landscape in dentistry.
While some might decry the loss of many independent dental practices, consolidation by DSOs allows for more advanced procurement and a deeper focus on efficiency. While individual dentists primarily focus on providing dental care, larger Dental Service Organizations (DSOs) have the advantage of employing specialized roles such as heads of strategy, Chief Technology Officers (CTOs), and other administrative functions. These professionals dedicate their time and expertise to actively identifying and evaluating opportunities to enhance their practices’ overall performance and efficiency. For young, high-growth technology companies, partnering with Dental Service Organizations (DSOs) presents an excellent opportunity to scale rapidly, bypassing the usual challenges of competing for a dentist’s limited time and attention.
Increased Consumer Demand:
According to the 2023 State of America’s Oral Health and Wellness Report, 92% of Americans recognize oral health as crucial to overall well-being. This shift, especially among younger populations, is propelling a demand surge for quality dental services. This growing consumer focus is reshaping the industry, leading to an increase in preventative care services and a higher standard of patient experience.
Additionally, the aging US population is bringing dental health into sharper focus, similar to other healthcare segments. Age-related oral health issues are escalating the need for specialized dental services. A GSK study illustrates some of the generational differences in oral health – younger consumers tend to be more diligent in their dental hygiene practices and have a better understanding of the link between oral health and other health complications. Over time, an increase in the desire for high-quality oral health is a benefit to dentists and the technology companies serving them. The landscape is shifting towards a future where superior dental care is not just a need but a priority for a growing and diversifying client base.
Staffing Challenges Prompting Efficiency:
One of the most pressing challenges faced by dental practices is administrative staffing. A third of dentists report needing more dental hygienists or assistants, and a quarter of dentists need help finding a sufficient number of office or administrative staff. Adding to this issue, dental assistants and hygienists frequently cite being overworked as a primary cause of job dissatisfaction.
This challenge is intricately linked to the need for more efficient operational practices. By integrating AI and automation technologies, dental offices can alleviate the administrative load on their staff and improve overall operational efficiency.
Investment Focus Areas at Ansa
We’re directing our investments toward areas where we believe technology can make a transformative impact:
- AI and Dental Imaging: AI is revolutionizing dental diagnostics. Beyond aiding dentists, AI technologies that have been trained on multitudes of dental images can help patients and insurers understand dental conditions more clearly. They not only support diagnoses and improve dental workflows but also open up new revenue streams by identifying additional treatment opportunities and allowing for better patient communication. The value of AI in dental imaging lies not in replacing dentists’ expertise. Instead, its strength resides in ensuring consistency across diagnoses, improving operational workflows, and elevating the quality of customer communication. Companies like Pearl, Videa, and Overjet are working to improve practice and dentist efficiency.
- Innovations in Business Operations:
Transforming Procurement: The current procurement landscape in dentistry is largely controlled by a few high-margin resellers, offering limited differentiation in their products. The time is right for innovators like Torch to challenge incumbents with stronger, more intuitive, and more automated procurement technology.
Customer Engagement for the Digital Age: Digital marketing and customer communication tools tailored for dental practices are changing how clinics contact and interact with their patients. These solutions are about attracting new patients and building lasting relationships, ensuring patient satisfaction, and ensuring lasting care. Platforms like Peerlogic are a game-changer for busy dental offices, where administrative staff are often overwhelmed, ensuring consistent and effective patient interaction without adding to the staff’s workload.
Modernization of Revenue Cycle Management (RCM) and Billing: The administrative side of dental practices, especially billing and reimbursements, is ripe for automation – particularly given advances in OCR and document ingestion. Traditional reimbursement workflows are typically labor-intensive and inefficient. Even modest improvements in these processes can yield significant time and cost savings. Automating these aspects of practice management with platforms like Retrace or Zentist not only streamlines operations but also allows dental staff to focus more on patient care and less on paperwork.
A Call to Action for Dental Tech Innovators
If you’re a founder or innovator in the dental technology space, we want to hear from you. And if you’re building in the space and raising a Series A-C, reach out to duncan@ansa.co or marco@ansa.co.
A Strategic Guide for Entrepreneurs Selling into Government Budgets
While there’s been lots of spilled ink on scaling a SaaS company, we’ve been surprise by the silence on scaling in defense tech. We’re breaking it with a strategic guide for founders. If you’re curious to know more, get in touch and we’ll send you the full guide. Thanks to our team, for helping drive our efforts in the space.
While government institutions worldwide are re-evaluating their military-industrial complexes to meet the demands of modern warfare, the defense tech sector is widening its scope to include novel categories that are now considered top military priorities. These indispensable sub-sectors include advancements in artificial intelligence (AI), robotics, autonomous systems, cybersecurity, computational capacities, human-machine interaction interfaces, biotechnology applications, and innovations in renewable energy & generation systems.
At Ansa, we actively invest in each of these sub-themes, from Series A through C, and are acutely aware of ‘what it takes’ to bridge the gap from seed to scale in defense tech.
Notably, these companies are not traditional SaaS businesses, so our firm has worked tirelessly to derive KPIs that can serve as leading indicators of what good looks like in building stronger dual-use companies and making world-class investment decisions. While many folks publish SaaS metrics (thank you, David Sacks, Alex Clayton, Jamin Ball, and dozens of others), few and far between have yet to do so on the new paradigm of defense tech.
We want to change that.
It’s no secret that one of the biggest barriers to entry in selling to the government, specifically the DoD, is their notoriously complex procurement and contracting processes. Before entering the defense market, it’s incumbent upon startups and dual-use vendors to evaluate whether the expected revenue generated from the DoD will outweigh the cost of acquiring and serving them. If the math isn’t mathing, then it’s all for moot. But if / when an initial DoD contract is landed, typically via an SBIR (Small Business Innovation Research) grant, there remains a challenging and unpredictable road ahead to transitioning into larger, longer-term revenue opportunities by becoming a program of record.
For most DoD-focused companies, becoming a program of record is the ultimate goal that lands them a line item in the DoD’s budget. It’s usually not apparent which companies will be the ones to become programs of record, but here are some of the signals we look for in companies pursuing this path and help influence the outcome.
SBIR Contract Conversion
- Many early-stage defense tech companies rely on SBIR grants for easier and quicker access to DoD contracts. Progressing from Phase I (concept development) to Phase II (prototype development) and eventually to Phase III (commercialization) is highly desirable.
- Phase I: Awards range from $50-250K over 6 months
- Phase II: Awards can go up to $2M over 18-24 months
- Phase III: No dollar value or duration limit
Diversification Across Agencies
- A sign of a scaling company is its ability to diversify its client base across multiple DoD agencies such as the Navy, Army, Air Force, and the Defense Information Systems Agency (DISA). We work with companies to leverage their success within one branch of service or funding agency and build a repeatable playbook that scales past founder-led sales.
ACV Collection
- Understanding the Annual Contract Value (ACV) collection as a percentage of the contracted ACV over the term of the contract is critical.
As companies mature, the gap between the contracted ACV and the actual collected ACV should narrow, marking a positive trend in business operations.
Beware of relying on vanity metrics, such as certain large IDIQ (Indefinite Delivery, Indefinite Quantity) awards, often described as hunting licenses for task orders that may or may not end up being funded as expected.
Ability to Find New Sources of Funding
- Unlike traditional enterprise contracts, DoD contracts do not always follow the concept of renewals. As certain existing contracts approach expiration, vendors often need to find new “pots of money” to continue funding commercialization efforts.
If you’re a Series A-C operator interested in seeing the rest of our guide, reach out to Hannah McQuaid, hannah@ansa.co, and Marco DeMeireles, marco@ansa.co.
Where Subscriber Lifetimes Outlive Funds
Hidden Opportunities in “Software-Driven Systems”
As an investor, I don’t know if I love anything more than backing a new set of physical products or experiences that integrate a subscription model to deliver value for consumers.
I gravitate towards businesses that broadly fall into two buckets: (1) the creation of a new market (eg. PlayStation Plus) or (2) the application of a new business model to improve an old market or service (e.g. Alarm.com, or Dadi):
While I am lucky to be a Board Member or institutional investor in many of these, I got involved because I deeply believe these businesses have some of the most compelling unit economics I’ve seen in my career. AND YET, they were underappreciated and ignored by investors for the majority of their early-stage lives, and, in some cases, through their IPOs!
But it is not often you see implied subscriber lifetimes outlast a fund’s life. When I do, I lean into it.
As a point of comparison, consider the overwhelming majority of mobile consumer subscription apps alongside the group above. Many a la carte applications have raised hundreds of millions in preference (individually!), but often these businesses have significantly worse fundamentals (e.g. churn, pricing power) and do not achieve enduring success.
The thread that ties the companies above together is that their subscriber base is centered around hardware or a physical experience. Importantly though, these are not businesses which simply slap a subscription on top of an existing offering (which we see a lot of in this market), but intentionally designed products that integrate software to create value across high frequency interactions or emotionally sensitive purchasing decisions.
And while investors have had an aversion to hardware-centric subscription models, it’s clear that not only can the retention dynamics be better than many segments of pure-play consumer apps, but so can the moats, purchase psychology, and long term FCF yield.
Many VCs don’t believe it and some are even skeptical after I walk them through the numbers like the above. I welcome that.
As Fred Wilson likes to say, this is how new funds outperform — avoid “me too” investing.
Lastly, I share this as an example of why I believe it is critically important for emerging investors to develop a nuanced thesis – ideally one refined through experience, and overlooked by the market. This is the only way I’ve found it’s possible to (a) gain the domain expertise needed for a sustainable investment advantage, (b) cultivate the open-mindedness required to see opportunity others won’t and (c) have conviction where others don’t.
When I look to the coming year, I am particularly interested in finding software driven systems within health & wellness, gaming and sustainability. I strongly believe the best businesses in these spaces will not be standalone mobile applications, but rather full-stack systems often with a physical or experiential component. I welcome the skeptics, just as we did in financing Dadi, Peloton and Latch, from Series A through IPO.
Founders who are building in this full-stack approach benefit from working with an investor who has walked the factory floors in Taiwan, knows complementary go-to-market channels, and has the product sensibilities in both hardware and software to drive increasing value for the customer each month.
Ultimately, customers are the voting machine, and they speak with retention. And as the data above suggests, an elegant “software-driven system” may be the best approach.
If that resonates with you, get in touch on Twitter.
Special thanks to @Joanne Schneider DeMeireles @Brett Bivens @Ian D’Silva @Nikhil Basu Trivedi and @Eric Crowley for helping crystalize my thoughts, the entrepreneurs I’m so fortunate to work with, and my partners over the years who enabled me to find my game and take smart risks in the businesses above. Originally published at https://marcodemeireles.substack.com.
Investing in “Confusion”
A key principle I’ve picked up from Jim Breyer over the years is to “invest in confusion.”
Massive confusion in growing markets represents an enormous arbitrage opportunity.
Confusion in this context can refer to a newly available technology, or to changes in the economics of supply, production and/or distribution of a good or service. It can also be driven by regulatory changes, changes in a competitive landscape and so much more.
While confusion manifests itself in different ways, it’s clear that not only are these the moments in time that are uniquely positive for emerging businesses, they are the cornerstone of my most successful investments.
This holds true across sectors. Consider:
- Databricks successfully commercializing Apache Spark in the face of legacy vendors with larger sales forces and larger spark investments (IBM)
- The ballooning impact of Facebook, Pinterest, Twitter, Reddit, et. al on our communities in contrast to Yahoo’s inability to move into social networks
- Netflix and Hulu’s clear victory delivering a streaming offering that captures, serves and delights subscribers before legacy media got there
I could share a list of 100 more examples, but you get the point. Massive confusion of legacy players often renders them unable to compete, and paves the way for a new set of enduring companies.
Often, the only option for these established businesses is to buy into the market, but they rarely pioneer a market when confusion runs high. This is partly why the NYSE did not build Coinbase, and Splunk did not build Confluent.
For investors, confusion presents an added benefit of making it less likely that the market is appropriately pricing related investments. Often, it is because these companies don’t fold into a simple existing total addressable market definition, and/or have elements (including business models) that not many have underwritten in the past.
The best way I’ve been able to navigate this is to (1) dig deep into the platform shift by speaking with every participant (customers, competitors, industry analysts, founders and operators) (2) actually listen intently – with “Dumbo Ears!” and (3) force yourself to tune out the crowd. I can’t tell you how many times I heard “No one makes money in media,” “Crypto is a fad,” or “Hardware is hard; it doesn’t scale” before backing teams proving folks wrong.
Do you believe there are certain shifts happening that established market leaders are confused about? Get in touch. I’d love to talk with you about it.
Why Now
A key question you hear in many investing conversations — but not enough!—is “why now?” Over the last 10 years, I’ve focused on becoming more disciplined in answering this question every time I map out a new market or explore a new business.
“Why now” could be that a newly-available technology enables a new service, such as the introduction of Broadband and subsequent roll-out of 5G, which catalyzed the widespread adoption of streaming-services like Netflix into homes and devices around the world. Or that patent expiry, like Invisalign’s, which cleared the way for Smile Direct Club, Candid, and Uniform Teeth to improve upon a dated approach with superior service and pricing. Or large behavioral shifts like the destigmatization of mental health, and related self-care movement, which paved the way for digital tools like Headspace to be a service consumers are proud to download and employers tout as a benefit rather than a fringe or shameful need.
Often, the simplest “Why Now” examples to understand are regulatory, which is what I’m going to focus on in this post. I’ll cover other angles in future posts.
To understand the power of getting the timing right, consider the following two examples:
US Consumer Finance: The Rise of Mobile Neobanks
- The Credit Card Act of 2009: Banned marketing of credit cards on college campuses, increasing the average age of credit card acquisition and leaving more young people in the last decade to depend on other means (ie. debit).Debit card users don’t want to overdraft, let alone incur related fees while working with a UX that left much to be desired. This interplay spurred a number of downstream consequences, including the increased adoption and engagement of personal financial management tools which can track spending in real time (among other functions).
- The Durbin Amendment of 2010: Reduced debit interchange rates for banks >$10B of assets from 155 bps + $0.04 to 5 bps + $0.21. This minimized the historically sizable profit pools that made servicing “low-value” checking accounts interesting to these types of banks. Thus, a legislative change disincentivized large incumbents from serving this population segment as they were precluded from charging sizeable fees to a large swath of accounts.What was the net result? Multiple segments of the market (young people, “underbanked” populations) began looking for better alternatives and upstarts like the Cash App, Chime, Varo Money, Moneylion, were able to cost-effectively take and better serve these customers.The timing of these two regulatory catalysts, coupled with a number of other factors, drove a strong “why now” for a large segment of fintech businesses. These businesses have seen strong product-market fit and billions of enterprise valuation creation.
US Telemedicine
Historically, the establishment of a “valid doctor-patient relationship” precluded the growth and penetration of telemedicine across the states.
Just 10 years ago, remote methods for establishing a valid relationship remained in legal limbo. Texas was the last US state to allow for remote validation following a landmark ruling involving Teladoc, and more recent regulation (ie Ryan Haight Act) has catalyzed the ability to facilitate online treatments and services across new use cases.
Now of course with the current pandemic climate, we are seeing even more widespread relaxing of these regulations which we can expect to pave the way for even greater use cases.
And this is what makes our job as investors and founders so interesting. It’s a constant exercise in understanding how changes to our world will enable greater innovation and usher in significant value creation. And how we capture that.
“Why now” unlocks this conversation.
Are you in the early stages of building something and have a compelling “why now?” If so, get in touch with me on Twitter. I’d love to talk with you about it.
LateralThinking.vc 