The AI Engineer’s Guide to Raising VC — Dani Grant (Jam), Chelcie Taylor (Notable Capital)

When should you raise venture capital: as early as possible, or when you need to scale? Dani Grant opens with that question, then makes it concrete with a show of hands: Do you need revenue to raise VC? Chelcie Taylor’s answer is no, provided you are approaching the right kind of fund.

The distinction changes what you need to prepare. A pre-revenue company is not automatically unready to fundraise; it may simply be a poor match for a fund whose investment criteria require operating metrics.

The next audience poll removes another apparent prerequisite: a finished product. At the earliest stages, Taylor looks at the founder, the team they can assemble, their recruiting ability and their vision. She expects the product at an eventual outcome seven to ten years later to look dramatically different from the original pitch. That makes the founder’s ability to recognize a market insight and adapt more consequential than having a beautiful first version.

Must you leave your job before raising? This answer is more dependent on the firm and partner. Some investors will consider a founder who has made progress on a side project and needs funding to make the leap. Taylor explains the underlying incentive: venture investors seek companies capable of returning their funds, so they value conviction. Leaving a job can signal that conviction, but it is not a universal admission requirement—and building a company solely to satisfy VCs is the wrong starting point.

Grant’s own experience makes the distinction tangible. Jam raised before it had a product, while her co-founder was still employed elsewhere. Their early conversations concerned what they were about to build and the prototype they planned to show. Grant found that potential easier to fundraise around than the much later task of producing a product impressive enough to make an investor immediately want in.

That leads to her recommendation to begin investor relationships before launching a formal raise. Asking for advice invites someone into the work while it is still taking shape; the earlier they participate, the more connected they can feel to its progress. The familiar advice-versus-money maxim captures that dynamic, but the practical action is simple: explain what you are about to attempt and ask for relevant input. On team structure, Taylor says co-founders are typical among the businesses her firm backs, while solo-founder exceptions exist.

What is the first step toward a pitch? The audience offers pitching, solving a problem and finding someone in an elevator. Grant brings the discussion back to access: a warm introduction is preferable, but many engineers do not know investors. Cold email remains an option because finding promising companies is part of an investor’s job. Taylor describes her own role as a kind of sales work: she is trying to get capital into the hands of excellent founders.

Taylor says that, when she asked colleagues about the cold emails they answer, those messages usually contained a warm signal: a relevant connection to something the investor was already thinking about. Taylor says she posts ideas on LinkedIn two or three times a week. Those posts are invitations to learn from people working directly in a market. She reports almost always replying and taking a call when a relevant builder responds with a request for a 15-minute exchange of ideas. Establishing that relationship early is easier than arriving for the first time with a fundraising process already underway.

The first real outreach example begins with a prior meeting and the question the founder asked there. It then gives a compact progress update: the co-founders are taking time off from their company, have built an MVP and have begun beta testing. The request is bounded—15–20 minutes to discuss go-to-market and, if appropriate, get pointers to other investors.

The audience identifies the small time commitment and personal context. Taylor adds the most useful property: she knows exactly how she can help. The founders understand that she is not a consumer investor and ask for pointers to someone who is. Before replying, she can already picture the conversation and prepare for it. Specificity removes the work of figuring out what a vague request actually means.

The email’s job is to earn a response. Grant recalls advice from one of her investors: “Don't think selling in email. Think be compelling.” Her own mistake was trying to put the entire pitch into the first message. Enough information to create interest is more useful than a long explanation of everything the company might become.

The next message links to a Notable blog post, connects its argument to what the founder is building and includes a GitHub repository. Taylor says the message arrived at seed stage; Notable followed the company for multiple years before leading its Series B. The repository gave her something concrete to inspect, while the response to the article established shared interests. Even when the business was not yet an investment fit, the founder was someone worth knowing.

Two further examples reduce the pitch even more:

• Share related thinking. A founder responds to an investor’s post with their own earlier writing on the same topic and asks for feedback. Taylor confirms that the relationship led to investment, despite the opening message saying nothing about the startup.
• Connect founding theses. Another founder explains that an investor’s observation is exactly why they started their company, and expresses enthusiasm for discussing the space with someone who has thought seriously about it.

Both messages make the likely conversation useful. They establish a shared baseline and offer an exchange of information, rather than requiring the investor to evaluate a complete company from an unsolicited message.

Once the email produces a meeting, what happens? Audience descriptions range from building a metaphorical Jenga tower to matching theses and getting to know one another. Grant proposes a role reversal: imagine you have a limited pool of money to invest in founders. What would make you choose one? An audience member supplies the central questions: Why you? Why now?

Engineers can over-focus on technology because it is the part they know best. Technical ability and a strong product still matter, but the early investment decision also depends on a distinctive insight into the market, distribution, product or customer segment. Taylor’s questions probe how the founder thinks and where that perspective comes from: lived experience, reading or observation. The goal is to make an unusual approach understandable and grounded, rather than simply assert that the technology is good.

A useful pitch starts with the core problem, then explains how the founder arrived at a particular solution. Taylor uses vertical voice AI as an example. The broad opportunity—automating phone-dependent work—is already familiar. The more revealing question is why a founder chose logistics and transportation instead of restaurants. Does that industry offer a workflow the product can become embedded in? Could the resulting data create a durable advantage? Those choices expose the reasoning beneath the product. A well-known problem can still support a compelling company if the founder identifies an opening they can address especially well.

The audience next raises return on investment. This exposes a second pitch mistake: explaining the present in detail without showing how it could grow into a billion-dollar company. At pre-seed, the useful question is often how future revenue could emerge, not how extensively a founder can describe a product or revenue stream that barely exists yet.

Taylor sketches a hypothetical answer built from buyer conversations: four prospective design partners have expressed interest, and those discussions suggest a possible range for annual contract value, or ACV. That is enough to begin reasoning about the next phase. Even without formal design partners, customer interviews and research can supply evidence of willingness to pay. Distinguish what customers have actually said from the revenue hypothesis it supports; together, they help an investor see a possible path to returning a meaningful portion of a fund.

Jam encountered this problem when its small MVP prompted investors to ask whether it was a product—or a company—rather than just a feature. Grant came to understand the question as a request for the expansion path. Her answer was: “This is the first feature of our future product.” She would then explain what that future product could become and how the company would get there. The small starting point became understandable once it had a place in a larger plan.

Would you invest in someone smart if you did not enjoy meeting them? Grant’s next audience poll draws few affirmative hands. The room’s ingredients for a good meeting include next steps, laughter, focus and a two-way conversation. Sometimes an investor’s question merely checks whether you have considered an issue. A long go-to-market monologue can overwhelm that purpose. Give the likely approach briefly, acknowledge that you have considered alternatives and offer to go deeper.

Taylor confirms that she often asks questions to understand a founder’s thinking. In a recent investment-committee meeting, a founder declined to invent an immediate answer and said he wanted to be thoughtful. The response impressed the investors because a persuasive answer is not necessarily a considered belief. Founders should know the basics of their own business, but they can pause on an unresolved question and ask what prompted it. Understanding the investor’s thinking can reveal what the conversation actually needs to address.

Pitching strangers under a perceived power imbalance takes practice. Taylor recommends putting less likely or less preferred fund matches earlier in the schedule. Use those meetings to become comfortable with the narrative and the questions it attracts. Then meet the more desirable partners—perhaps a pre-seed investor already interested in your category and in AI engineers—with some repetitions behind you.

The expected questions form a practical preparation sequence:

1. Problem and timing: What are you solving, why are you suited to solve it, and why now?
2. Solution and iteration: What could the product look like, and how will you maintain development velocity while learning what works? A coherent explanation can precede both a product and mock-ups.
3. Customer: What is your ideal customer profile, or ICP? Two or three candidate profiles are acceptable if you can explain how you will test them and narrow your focus.
4. Monetization: What do customer or design-partner conversations suggest about willingness to pay, price range and the basis for charging?

The point is to arrive with a reasoned position and a way to learn, rather than pretend every decision is settled.

Discussing these alternatives gives an investor something specific to respond to. Taylor describes bringing in examples of similar companies and their deal sizes, helping the founder refine a possible revenue trajectory.

Recruiting capacity is part of the investment case. Taylor calls the team probably the most important early-stage consideration. Expect questions about the first people you would hire and how you would attract them. Her example is a founder who knows five engineers at a particular company who might want to join. The useful evidence is a credible recruiting network and the ability to build momentum; those people are prospective hires, not a team already secured.

Competition is a question engineers may have less experience answering in their day jobs. Start by acknowledging competitors, including adjacent products that do not do exactly the same thing. Then map the market along meaningful dimensions and explain the opening you intend to occupy. Claiming that no competitors exist makes it harder for an investor to trust your understanding of the market.

Taylor’s example begins with ZoomInfo’s established role in data enrichment. She then characterizes Clay as entering with a thesis of aggregating the most data sources, and points to its fundraising and momentum as evidence that the positioning was compelling. The lesson is the structure of the answer: identify the existing category, describe a different approach and explain where your company sits. The aggregation superlative is Taylor’s characterization of that thesis.

Investors focused on a category may already have spoken with the companies you are discussing. Hiding them risks turning a positioning question into a trust problem. Acknowledge the activity in the space and explain how your approach differs. When the conversation permits, reverse the question: what did the investor like about another company, or why did they pass? Taylor welcomes that kind of exchange because it helps the founder understand her reasoning.

For an early go-to-market answer, bring a directional plan grounded in comparable companies. Taylor initially mentions usage-based pricing, then distinguishes monetization from a product-led growth, or PLG, approach to reaching and growing customers. Explain which comparable businesses inform your plan, why their approach seems relevant and what initial pricing tiers might look like. Ask what the investor has seen work with PLG companies. Pricing and distribution still need experiments; the pitch should show how you are thinking about them.

Choose technical depth based on the person across the table. Research whether the investor is highly technical; otherwise, an initial pitch that spends all its time on implementation can leave no room for the founder, product vision, go-to-market or future business. Taylor says her firm sets aside a separate one-hour session with its VP of technology for technical diligence. That separation gives the first conversation room to establish why the company is worth investigating.

The meeting is also your opportunity to decide whether you want this investor involved. Taylor respects founders who can explain what they need to reach the next phase and what they are looking for in a partner. Define those needs so you can evaluate the investor’s fit while they evaluate yours.

As the meeting ends, ask about next steps if the investor has not offered them. Get a concrete explanation of the firm’s process and what to expect over the coming week, two weeks or month. Have materials ready to send immediately. A pre-seed or seed raise does not necessarily need a full later-stage data room, but the deck should cover the business questions discussed, how much you are raising and what the capital will fund.

Grant points to Jam’s original 2020 fundraising deck at deck.jam.dev, also identified in Jam’s public seed-deck announcement. She says it remained online because they had never removed its Cloudflare DNS rule. Her advice is to save the slides for the follow-up: screen-sharing a deck can send people into passive webinar mode, while the meeting itself benefits from conversation. The deck becomes a brochure for the investor’s teammates.

Whether you meet a general partner who can write a check or a more junior associate, the person across the table must still explain to colleagues why capital should go to your company. Send a compelling deck together with a short, forwardable blurb. Those materials help your contact advocate internally and can also travel to other seed or pre-seed investors. Taylor closes by making the outreach advice actionable herself: she invites founders to contact her by email.