Founder vs Co-founder vs CEO and Founder: Founder is the origin role, the person who started the company. [Co-founder] is the same role when there's more than one person at the origin (a co-founder is just a founder with company). [CEO and Founder] is the title combination, a founder who also currently holds the CEO job. The cap table and certificate of incorporation determine who's a founder; current employment determines who's a CEO; the two questions are independent.
A founder is a person who started a company by originating the idea, building the first version, and taking the early risk before others joined. It is the company's origin role and, unlike most roles, it does not end when the person stops working there. Once a foun...
Corporate bylaws are the internal operating rules of a corporation. They govern how the company makes decisions, holds meetings, elects directors and officers, conducts shareholder votes, and handles routine corporate actions. The document is adopted at incorporation (typically at the initial organizational meeting of the board), and amendable by board or shareholder action depending on the bylaws themselves. Unlike articles of incorporation, which are filed publicly with the state, bylaws are an internal document kept in the corporate records, addressing offices, shareholder and board meetings, officer roles, stock issuance, indemnification, and amendment procedures.
The major sections of a typical set of corporate bylaws: offices a...
A warrant is the security granting the holder the right to purchase shares at a fixed exercise price for a defined period. Similar in mechanics to a stock option but issued to non-employees (banks, lenders, strategic partners, advisors), warrants typically run 5-10 years and are commonly attached to venture debt deals, strategic partnership agreements, and bank facilities. It is the structural analog of a stock option for non-employee parties and a frequent component of late-stage financing structures.
The warrant mechanic:
Token economics is the discipline of understanding and modeling the per-token costs and revenue of AI applications. Tokens (sub-word units of text) are the unit of pricing for most LLM APIs and the basis on which AI application unit economics must be modeled. The model includes input tokens (prompt + context + RAG content), output tokens (model response), cache savings, and the per-query economics that determine whether AI applications are profitable. It's the financial layer beneath every AI application.
What a token is:
Token: sub-word unit of text used by LLMs. Roughly 0.75 English words per token, or 4 characters.
Tokenization examples:
A patent is a government-granted intellectual property right giving the inventor exclusive rights to make, use, sell, and license an invention for a limited period. The term is typically 20 years from filing for utility patents, in exchange for publicly disclosing how the invention works in sufficient detail that someone skilled in the field could reproduce it. Patents are issued by the US Patent and Trademark Office (USPTO) at the federal level and have no state-level equivalent.
The three patent types:
MRR (Monthly Recurring Revenue) is the normalized monthly value of a subscription business's recurring revenue at a point in time. It's calculated as the sum of all monthly-normalized subscription contracts: a monthly subscription contributes its full monthly fee; an annual contract is divided by 12 to get its monthly contribution. MRR is used heavily by month-billed SaaS companies and consumer subscription businesses where monthly granularity matters for operating decisions. It is mathematically equivalent to ARR (MRR × 12 = ARR) but with different operational implications because monthly tracking captures shorter-cycle business dynamics. It's the alternative to ARR most useful at consumer subscription, SMB SaaS, and month-to-month bus...
An advisor is an outside expert who provides part-time strategic guidance, introductions, or domain expertise to a startup in exchange for equity. The relationship is formalized in a short advisor agreement and distinct from board members (advisors have no fiduciary duty and no voting power), from investors (advisors are not buying equity), and from consultants (advisors are ongoing relationships paid in equity, not project work paid in cash). The role is the most common way founders extend their leadership reach in the first three to four years of a company.
The categories that matter: technical advisors (a senior engineer or domain expert who reviews architecture decisions or vouches for the founders to investors, typically 1 to 4...
An earnout is a contingent acquisition payment tied to the acquired company hitting post-close milestones over a defined performance period, typically 1 to 3 years. Milestones cover revenue, EBITDA, product launches, customer-retention thresholds, or other operating metrics, and the structure is used to bridge valuation gaps between buyer and seller when the buyer doesn't want to pay up front for value that depends on future performance. It is one of the most-negotiated and least-loved acquisition mechanics, because it transfers performance risk from buyer to seller and gives the seller limited control over the metrics they're now paid to hit.
The typical structure: an earnout represents 10 to 40 percent of total deal value (sometim...
Retrieval-Augmented Generation (RAG) is the AI pattern of pulling relevant context from a knowledge base and including it in the LLM prompt at inference time. Retrieval typically uses vector search. RAG allows the model to answer questions or perform tasks using information it wasn't trained on, or that may be more recent than its training data. RAG has been the dominant pattern for knowledge-grounded AI applications since 2023. It's the bridge between general-purpose LLMs and specific organizational knowledge.
How RAG works (the pipeline):
A design sprint is a five-day process for solving big product problems through ideation, rapid prototyping, and real-user testing in a single compressed week. It was developed at Google Ventures by Jake Knapp and is designed to bypass the usual months of debate and produce a validated direction before the company commits to building. It was popularized through Knapp's book Sprint (2016, co-authored with John Zeratsky and Braden Kowitz) and became one of the most-adopted structured product processes in the late 2010s.
The classical five-day schedule: Monday, understand (map the problem, interview experts, choose a focus area), Tuesday, sketch (each participant generates solutions individually, leading to a detailed solution ske...