You have plenty of experience. Maybe you are over 40, or fresh out of college, and you realize that there is no tech job out there. More precisely, 1000 applicants for any job opening. And even if you land a job, the risk of layoff within a year is high. It is to the point that launching a startup may seem safer and less stressful. But how to start? What are the challenges? How to maximize the chance of success? In this article, I share my experience launching an AI startup, including fundraising. My perspective is that of a founder in US, with co-founder outside US. Also, not my first startup: I sold the previous one in 2020 and was VC-funded in 2008. I also succeeded on the real estate market.
Before you start
Don’t leave your current job if you have one. It’s far easier to start when you are still employed. Try to automate your job as much as you can, to free time. In my last corporate jobs, I achieved extensive automation and was working for the company about 2 hours a day. As a BI manager, my reports and forecasts were automatically produced and emailed to stakeholders. Of course, I learned how to do it, but I also benefited from a very good relationship with the IT team. They allowed me to directly access the databases with my own code rather than painfully slow, non-automated dashboards. And run cron jobs for automation (to send my email featuring insights and alerts).
With plenty of free time on your hands, you can take on another job, say consulting, to further boost your revenue to jump-start your new venture, while still full-time on W2. Consider buying a modest home if you are renting and interest rates make sense, as it will be more difficult after. And that’s an asset you can leverage, especially if your buy when everyone sells.
Build trust and credibility
Start open source. Publish a lot of very innovative content. Maybe write a book with a well-known publisher. Be a speaker at reputable conferences that target stakeholders rather than academics. Grow a large network. You may even use it to generate advertising revenue as an AI influencer. There are many around, but very few with deep tech knowledge. In my case, my Gen AI newsletter has 200k subscribers; Nvidia is one of my partners. In the end, I will use my newsletter for my startup, thus reducing advertising expenses. Also, comment in relevant forums. Not shallow content, but content that shows considerable maturity, not AI generated. Have a strong portfolio, including on GitHub.
Hire a CEO, sales executive, and team members
In my case, I lack sales acumen, I am an introvert, not charismatic, and I don’t see myself as a CEO. This is true for many tech entrepreneurs. What I bring to the table is vision, credibility, and building the future of AI. Not just another new agent.
It was not my plan to start another startup. But when my contact, global AI director at Fortune 100 reached out to me to launch a company based on the AI technology I developed for them, I could not resist. I named him CEO without hesitation. The timing was perfect. I then brought in an executive with success selling technology in the US to large companies. One who worked with me in the past. I was lucky that the timing was perfect for him. In both cases, I don’t have to shell out $500k in salaries. It is incremental based on deliverables, starting at zero with reasonable equity in the venture for each.
As for team members, I selected the most qualified that reached out to me. Not using recruiters or agencies. Not even job ads. All remote working from India, Brazil, and Australia while I am in Seattle. Paid with a combination of cash and equity. Saving on agency fees (thus able to offer better rate) and without offices. And co-founder and other executive partners brought in great hires that they are connected with. In SEC filing documents, my Delaware corporation has my main home address listed as headquarter. We don’t have offices, everyone is remote. All this, including oversees staff and CEO, results in a very low burn rate, about 10 times smaller than my peer competitors. To the point that I can do without funding other than my own if I wanted to.
Minimizing costs, unusual revenue streams
Be 100% online, with no office. Hire abroad on your own without HR. My experience with recruiters is that they pretend to know my technology, but they don’t and send irrelevant, expensive candidates. Ask your staff to use AI as much as possible to automate work such as coding. Do not hire LeetCode geniuses or folks with too much experience with tools you are competing against: they’ll want to use those tools rather than understanding and building your own. Because that’s what they feel comfortable with.
Optimize your tax situation, perhaps with a small LLC (shell company) in addition to your startup. Have a tax address in a place that is business friendly. Minimize legal costs: use AI to produce legal documents when possible and use common sense to assess when you really need a lawyer. Avoid coaches and incubators asking upfront fees to find investors or clients. If they knew how to get them, they would create their own startups instead. Likewise, conferences asking $10k to participate in their “alpha” program to meet investors, are mostly scams. Legit scams, but scams, nevertheless. In addition, wasting time that you could spend instead on developing your product or reaching out directly to potential clients or investors. Do not spend too much time on patents; instead file provisional patents yourself and don’t spend more than 3% of your budget on that.
Eliminate all expenses. I don’t have a car, I don’t fly, I spend zero on health care and don’t have health insurance (please do get one, my situation is unusual). I self-learn so don’t spend money on training. Anyway, much of the available training is outdated and boringly slow. But you need to know what your competitors are up to, precisely understand what they do, and what makes you different. Look at their executive profiles and connections.
To generate revenue, I rent out one of my homes. If you have less money, maybe you can have roommates. I also sell books, advertising, have passive income from investments, and write paid articles. You could add teaching your own classes, even corporate teaching (can be future customers), as long as it does not take too much of your time. Or selling some assets like a blog or some IP. Finally, I don’t take vacations because I love what I do. If you feel that operating your startup is like working, you are on the wrong boat.
Connections, connections, connections
Unless your startup sells nothing — for instance if it focuses on arbitraging the stock market for its founders exclusively — you need to build a strong network. And partner with co-founders who are well connected.
The first rule is to connect with people active on LinkedIn to increase the chance of a response. Identify their needs and offer for free something that is valuable to them, for instance a free copy of your book. Never try to sell anything in an introductory message. Likewise, identify people, no matter how well-know they are, who try to get free stuff from you. Don’t take the bait. These could be VCs interested in getting free advice from you, possibly detrimental to your startup, on the false pretense that they can fund you. Ask connections to get you connected to relevant people they are connected to, offering the same in return.
Raising money
Unless you have an MVP, an active automated platform, and recurrent revenue from customers, at least $100k per year, VCs won’t talk to you except for the wrong reasons: gaining insights about your business because they are investing in a competitor. Even then, there is no guarantee that you will get any money from them unless you are a close friend or properly introduced. Again, don’t pay fees for introductions. If I learn that someone paid money to get introduced to me hoping I will invest in them, it’s a red flag: a sign of a founder burning money inefficiently. I would stay away from them.
It is best to assume that you won’t get VC money for 2-3 years, maybe more, and plan accordingly. In my previous company, I never received funding and run it out on my own for 8 years — very profitable, large clients — till the end (acquisition). Thanks to that, I earned a hard 50% on the payout. I’ve heard stories of founders selling for 5 times more than me (above $40 million) yet ending up with less than me due to dilution and VC contracts with terms that were very unfavorable to founders.
Angel money should be your first route at the beginning. Better, self-funded. However external funding can act as validation for your business and be more valuable than the cash itself. Another interesting story is that VCs seem to like if you pay yourself some fair compensation to operate your business. I don’t, to extend runway and better compete, but my situation is different as I am an investor myself.
Also, look for clients and investors abroad. These days, they are becoming attractive compared to US. And wait for investors to contact you rather than the other way around. However, beware of investors contacting you for the wrong reasons, and fake investors. The latter are easy to detect; they’ll ask you to pay an upfront fee.
Conclusions
Startups are for people who strive in adversity and chaos. For people who quickly learn from setbacks, understand that personal revenue is not a smooth predictable path, but one with ups and downs eventually leading to great financial security if properly orchestrated. Key factors for success: adaptability, patience, resilience, quick bounce back, planting multiple seeds while staying focused, being street smart, leveraging all sorts of tricks (business hacks), helping luck happens, see it and seize it when it happens, saying no to time and money wasters, nurturing connections, working smart not hard (automate), making mistakes, determination within reason (know when to make a U-turn), a strong dose of common sense, derisking, and not burning bridges. For skills that you don’t have, partner with a complimentary co-founder who has them. And listen to advice in areas where you don’t shine.
Age should not be a barrier. You can partner with someone younger. And despite my age, I take a lot more risks and have fewer medical issues than many people 40 years younger (some medical problems may be linked to living in constant stress, such as toxic job or boss). The kind of risks I take even scare some people. And there are some that I don’t take, and mistakes I don’t make, red lines that I don’t cross. It’s all about risk intelligence, which probably explains why I tend to experience fewer problems compared to others. The ability to predict the consequences of your actions and how it can backfire (to assess if it worth doing it), also goes a long way.
About the Author
Vincent Granville is a pioneering GenAI scientist, co-founder at BondingAI.io, the LLM 2.0 platform for hallucination-free, secure, in-house, lightning-fast Enterprise AI at scale with zero weight and no GPU. He is also author (Elsevier, Wiley), publisher, and successful entrepreneur with multi-million-dollar exit. Vincent’s past corporate experience includes Visa, Wells Fargo, eBay, NBC, Microsoft, and CNET. He completed a post-doc in computational statistics at University of Cambridge.
