Matt Weinberg | Max Ventures
Venture Wishlist shares ideas and themes VCs want to fund. Brought to you by Purpose Built venture studio.
Matt Weinberg is a Partner at Max Ventures, an early-stage venture firm focused on day zero investments. The firm currently has approximately $100 million AUM across three funds, making 80+ investments, with a company incubation arm that has launched 8 companies. Max Ventures is led by Matt and his partner Ryan Darnell.
Matt’s Venture Wishlist
💼 Intuitive Marketplaces: Interest in conversational AI interfaces that help users navigate complex decisions through preference-based recommendations, particularly in areas like fashion and travel.
🏛️ GovTech Innovation: Interest in tools that streamline government processes, including FOIA request management systems, factoring marketplaces for government contractors, and better data infrastructure for government agencies.
🔄 AI-Enabled Roll-Ups: Opportunities in consolidating fragmented, service-oriented industries and applying AI to create efficiency and value, with particular interest in government vendors and service providers.
🛡️ Novel Insurance Models: New insurance products, such as protection against sophisticated scams targeting the elderly, particularly when combined with preventative technology.
🔎 Research & Advisory Marketplaces: Solutions that help buyers navigate the increasing noise in vendor landscapes, providing authoritative research and à la carte data points instead of expensive comprehensive reports.
Other insights:
🌐Rise of Entrepreneurship: AI is enabling a new wave of entrepreneurship by reducing barriers to entry, allowing non-technical people to build technology solutions more easily.
🏢 Government Data Monetization: Potential for government agencies to better leverage and potentially monetize valuable datasets they generate, particularly for large corporate users.
Full Interview
Tell me about Max Ventures.
Max Ventures invests at the earliest possible stages, often pre-seed and pre-product. We pride ourselves on being a day zero fund, meeting talented entrepreneurs when they're considering what to do next or just after launching. Our third fund closed a few months ago, bringing us to approximately $100 million in AUM across all funds.
What industries do you focus on?
We're primarily a generalist fund, but about a third of our investments are in healthcare, where we've established a strong track record. We're also developing a focus in insurance after incubating a company in that space, which gave us deep industry insights.
How does your incubation strategy work?
We've built eight companies over the past seven years. Our approach involves identifying market problems, developing solutions with venture-scale potential, and partnering with exceptional talent. Unlike many studio models, we structure our cap tables fairly, recognizing that founders will be working on these businesses for a decade while we're involved for about a year. Beyond direct returns, the incubation process introduces us to a proprietary network of talent and opportunities we wouldn't encounter through investing alone.
What technological trends are you tracking?
AI is creating opportunities across multiple verticals. For consumers, we're seeing potential for intuitive marketplaces that use AI to understand user preferences in fashion, travel, and other areas. For example, what if you could go to a fashion website and upload images of what you like, and get recommended items you’d actually want to buy.
"Intuitive marketplace"? Is that like talking to AI for purchasing advice?
Yes, it's having a conversational interface where I can express my preferences about travel, fashion, etc. The AI would respond adaptively - if I mention disliking white dots on clothing, it acknowledges and avoids patterns. Or if I say I don't like hotels near noisy main streets, it surfaces quieter options that match my aesthetic preferences.
So it's adding search parameters not built into the original site, remembering preferences, and searching across different dimensions.
Exactly. The question is whether this becomes one big system or if there's room for different approaches.
And does the data live on the seller side or buyer side? I've been wanting a way to carry my data when interacting with LLMs without giving them everything - perhaps using privacy-preserving methods like homomorphic encryption.
Maybe there's a middle ground where platforms like Etsy, Expedia, or Mr. Porter keep proprietary datasets but build better tools around them. They could ask 5-6 orientation questions when you visit, let you upload outfits/brands you like, and then remember you. It's not as seamless as a fully portable solution, but it keeps data siloed if that's preferred.
Interesting.
More broadly, AI is democratizing entrepreneurship by making technology development accessible to people without traditional coding skills. You can bring products to market faster, which will result in an explosion of entrepreneurship.
As a former Obama White House official, there are trends in government that I like.
Like what?
I think there's an opportunity for a factoring marketplace for government contracting. The idea was inspired by a similar concept proposed for healthcare between providers and payers.
In situations where there's a significant time lag between invoice submission and payment, factoring becomes valuable. While healthcare might be challenging to underwrite due to claims denials and rejections, government contracting is particularly promising because the government almost always pays (essentially "triple A" reliability, the payment delays stem from bureaucratic processes, not payment intention, and the current payment process is extremely manual requiring multiple signatures and approvals.)
A factoring marketplace would be especially valuable for smaller government contractors (not the Lockheed Martins of the world). These long-tail contractors face inconsistent payment timelines despite the government's eventual reliability in paying. This kind of marketplace doesn't currently exist but could benefit both contractors and potentially government efficiency.
Why doesn’t that exist already?
Some analog versions of government contract factoring do exist, particularly for specific programs like SBIR grants, where banks will bridge funding gaps between winning and receiving payment.
To properly underwrite government contracts at scale, you'd need a platform that ingests and evaluates all contracts across the diverse government ecosystem (federal, state, county, municipal) and sectors (defense, energy, education, health services).
I haven't seen a broad, holistic digital marketplace tackling this, or even one focusing on a specific sub-sector. Building lending marketplaces is challenging - you need upfront capital, likely need to fund initial deals yourself, and require backing from major lenders like BlackRock.
Beyond factoring marketplaces, there's growing interest in "AI roll-ups" - where investors buy outdated service-oriented companies and apply AI solutions to create value. VCs are funding companies pursuing roll-up strategies in analog, service-heavy industries with long tails of small businesses or sole proprietors (accounting, insurance sales, etc.) that could be transformed with AI. Some new funds focus exclusively on this thesis.
What are some of the opportunities you’re seeing that are overlooked that could definitely be enhanced with AI?
One specific area I'm exploring is FOIA (Freedom of Information Act) request processing. Having worked in government, I know FOIA requests are extremely painful and time-consuming, requiring extensive manual work from compliance and legal teams.
While some software solutions and consultants already serve this need, there's potential for an elegant AI solution that could ingest data more efficiently, generate more accurate FOIA responses, and pull information from different departments automatically. This solution would be highly transferable across different cities and counties because the fundamental requirements are similar enough to make the product exportable to multiple government entities.
Currently, governments don't commercialize or monetize their valuable data assets. For example, when the Department of Energy or NOAA funds climate research on specific watersheds, the resulting data is incredibly valuable to energy companies, insurers, and reinsurers. Yet this data is often made available as open source.
Similarly, transportation data and public health data are being leveraged by private companies like Google and Uber without governments receiving compensation. There could be a way for governments to monetize this data, perhaps with a tiered approach where individuals and small businesses can access it for free, but larger corporations would pay.
GPS is another example of government data that has generated tremendous value in the private sector.
Weather prediction as an industry has been built almost entirely on government data. Despite this, it seems policy decisions have historically been intentional about not charging for access to these valuable data resources.
There could be a tiered approach where government charges corporations for data access while keeping it free for individuals, with the revenue improving data infrastructure for everyone. Current OpenGov data sites, even the best ones, are difficult to navigate.
This improved data infrastructure could help smaller cities, counties, and less sophisticated states enhance their data capabilities. One ambitious idea would be creating a government-side "EHR" (Electronic Health Record) for citizens. With your Social Security number and tax return data properly integrated, the government should automatically know your eligibility for food stamps, housing vouchers, and other benefits.
Currently, data is scattered across different government agencies in isolated pools. Creating a comprehensive data warehouse for each individual would provide tremendous value both for citizens and for reducing government bureaucracy.
Many other countries automatically file taxes for citizens. If you're a W2 employee with an uncomplicated situation, the government already knows what you earned and could calculate your taxes, giving you a chance to review for errors. This approach has been blocked in the US by two main constituents: tax preparation companies who would lose business, and anti-tax advocates who prefer that filing remains painful because they believe it maintains resistance to income taxes.
Empowering government with technology faces resistance from special interest groups that prefer the status quo. There's an emerging issue with new technologies making it easier to apply for government benefits, grants, and contracts - the government lacks the technology to efficiently evaluate the increased volume of applications, creating potential bottlenecks.
If we're creating tools that make application submission easier, we should simultaneously develop technology to help government process this influx. Many startups aren't considering both sides of this equation.
Regarding other opportunities, vertical-specific AI agents show promise, particularly in compliance - a manual, costly process across healthcare, government, and insurance industries. An AI solution here would be elegant.
The insurance industry, especially reinsurance, presents opportunities through live data and updated underwriting models. This extends to non-obvious sectors like shipping, transportation, and construction.
A concerning area is elder scam protection insurance. Modern scams are becoming dangerously sophisticated - for instance, using voice harvesting from robocalls to create manipulated audio that could impersonate someone to their elderly relatives.
Scammers could use voice manipulation to request funds from elderly parents, saying something like "Hey mom, can you transfer me some funds?" I wonder if there's an opportunity to create a specialized insurance product that would protect elderly people from such scams. This could cover situations where seniors transfer money from their accounts due to fraud, especially in cases where banks refuse liability.
It could be interesting to combine an insurance product with software that helps defend against scams.
My family and I have created a safe word system to protect against voice scams. Unless the caller uses our predetermined safe word, we know it's a scam. We were careful not to discuss this safe word near phones or recording devices - we went for a walk in the woods to establish it. Now if anyone contacts a family member claiming to be one of us but doesn't use the safe word, we immediately know it's fraudulent.
Any underserved areas you find interesting?
As AI reduces content creation costs to nearly zero, the marketing noise in B2B will increase dramatically. Two ways to combat this are authoritative research and word-of-mouth recommendations. Two of our portfolio companies reflect this thesis, Elion which is the premier expert research marketplace for healthcare technology and Noble, which is becoming the ubiquitous word-of-mouth or warm-referral option for people making software decisions.
For less complex decisions like choosing a CRM or hotel, talking with previous users is invaluable. Could this approach work for government buyers overwhelmed by RFP responses?
Research and alternative data present another opportunity. Currently, when I need specific stats (like "aviation market in Colombia"), I encounter research marketplaces charging $5,000 for entire reports when I only need one data point. I've never been targeted by Gartner with an ad offering just the stat I need. I'd gladly pay $100 for a single data point rather than $5,000 for a full report.
There's an opportunity to disaggregate research reports into a la carte products - like Getty Images but for data points, charts, graphs, or specific paragraphs from research reports.
That's interesting. General market stats do show up in SEO, but not narrow stats that follow you around.
Google performance ads exist, but I've never been targeted on Instagram with Gartner offering a graph of exactly what I searched for. Market data is either inaccessible ($5,000 reports), untrustworthy (random Google image search), or occasionally in trusted sources like the Wall Street Journal.
This interview has been edited for length and clarity.