Parties and effective date
Legal names, addresses, signing authority, and the exact date the agreement starts and ends.
How-to — — by Mahmoud Zalt
Learn how to write contracts without a lawyer: when DIY is safe, the must-have clauses, AI drafting tricks, redlining steps, and a safe signature workflow.
The honest test is the blast radius of the worst plausible outcome. If a deal goes wrong and the worst case is a few thousand dollars and a slightly awkward email, a DIY contract is almost always fine and the lawyer cost would have eaten the upside. If the worst case is litigation, regulator attention, a chunk of your cap table, or a personal liability claim, you are in lawyer territory and should treat the DIY shortcut as false economy. Most founders never need a custom drafted contract for short freelance gigs, small monthly retainers, simple NDAs, or basic service agreements with consultants. They do need a lawyer the first time they hire an employee, take outside money, sign a multi year vendor commitment, or operate in a regulated space.
| Dimension | Traditional | With Sista |
|---|---|---|
| Freelance gig under one month | Simple SOW or short services agreement, signed digitally | Multi year retainer with exclusivity or non compete clauses |
| Vendor or SaaS reseller deal | Standard order form with basic limits of liability | Custom MSA, data processing addendum, or revenue share |
| Hiring help | Contractor agreement for a clear project under a few thousand dollars | Employee offers, equity grants, or anything with payroll obligations |
| Confidentiality | Mutual NDA for an intro chat with another founder | NDA covering a real M&A, fundraising, or licensing conversation |
| Regulated work | No regulated data, no money movement, no minors involved | Healthcare, financial advice, kids products, EU consumer terms |
A contract is just an agreement written down clearly enough that a tired stranger can read it later and tell who owes what. Most DIY contracts fail not because they are missing fancy clauses, but because the boring parts are vague: who exactly is signing, what exactly is being delivered, on what calendar, at what price, paid how, and what happens if either side wants out. If those six things are crisp, the contract is doing its job for the vast majority of small business deals. Lawyers add value on the edges (limits of liability, indemnity, dispute resolution, governing law), but the spine is the same for a freelance brief and a Fortune 500 SOW: parties, scope, price, payment, IP, exit.
Legal names, addresses, signing authority, and the exact date the agreement starts and ends.
Concrete deliverables, what is explicitly out of scope, and how change requests are handled.
Total fee, schedule, late fee policy, currency, and which side covers taxes or platform charges.
Who owns the work, when ownership transfers, and what either side may reuse afterwards.
How either party can exit, notice period, and what happens to deliverables and payments at exit.
What stays private, for how long, and which country or state law settles disagreements.
Yes, for everyday business contracts, modern AI can produce a first draft that is cleaner than most templates floating around the internet, and it can do it in the time it takes to brief a junior paralegal. The trick is that AI is excellent at structure, language, and consistency but it does not know your jurisdiction, your prior agreements, or the specific risk tolerance of your business. Used as a drafting partner inside a clear workflow, it removes the blank page problem and forces the boring sections to actually exist. Used as a one shot oracle (paste a prompt, sign the output), it is a real liability because the output reads confidently even when a clause is wrong for your country, your industry, or your counterparty.
Produces a complete first draft with all standard sections instead of a half remembered template.
Turns a short brief into a clean draft in minutes, not days, and iterates on changes instantly.
Keeps terminology, defined terms, and numbering aligned across long documents.
Rewrites legalese into clauses a non lawyer counterparty can actually read and sign.
Does not reliably know what is enforceable in your country, state, or industry.
Will produce a clause that sounds correct but is wrong for your specific facts; needs review.
The pattern I run on my own business is simple. An AI Employee in Sistava holds the brief, the past versions, the standard clauses I have already approved, and the counterparty notes from previous calls. When a new deal comes in, the employee drafts version one, I redline in plain English, the employee produces version two, and only the final draft goes to a lawyer for a short review if the deal warrants it. This collapses a one week cycle into an afternoon and keeps the lawyer focused on the actual risk questions instead of formatting and typos.
The bigger unlock is that contracts stop being a one off chore and start being a system. Your AI Employee remembers the standard NDA you use with intro chats, the freelance template that worked last quarter, and the vendor terms you push back on every time. New deal lands, you describe it in two sentences, the right draft appears with your usual clauses already populated. That is the difference between writing contracts and operating contracts: same words, completely different cost per deal.
Redlining sounds intimidating because it has its own vocabulary, but the actual job is asking five questions in order and pushing back where the answers favor the counterparty too hard. The standard playbook for a small business reviewing a vendor MSA is to read the liability cap first, the termination clause second, the IP and data clauses third, the payment terms fourth, and the dispute resolution last. If any of those five is materially worse than industry norm, you mark it, propose a swap, and send it back. You do not need to know contract law to do this well, you need to know what good looks like, and AI is genuinely useful for showing you that benchmark in plain English.
A safe DIY workflow is not about avoiding lawyers entirely, it is about routing the right work to the right tool so the lawyer (if you use one) sees a clean draft and the deal moves fast. The trap most founders fall into is mixing drafting, negotiation, and signing in one chaotic email thread, then losing track of which version anyone actually agreed to. A clean process separates each step, keeps a single source of truth for the latest draft, and uses a real e signature tool at the end so both sides have a tamper proof record. Run this workflow every time and your contract operations stop being a source of risk and start being a quiet background asset.
Yes, in most jurisdictions a contract is binding because of mutual consent, clear terms, and consideration, not because of who or what wrote the draft. An AI drafted contract that is reviewed, signed, and exchanged the same way as any other contract is enforceable in the same way. The risk is not in the drafting tool, it is in not reviewing the output carefully before signing.
Not for one off small gigs. A short statement of work plus an invoice with payment terms is enough for most freelance projects under a few thousand dollars. An MSA pays off once you have a repeat client where you sign multiple SOWs a year, because the master agreement holds the boring clauses (IP, confidentiality, payment, termination) and each SOW just references it.
Yes, and it is one of the highest leverage uses. Paste the counterparty contract, ask the AI to flag clauses that deviate from market norm for a small business, and ask for plain English alternatives. Treat the output as a checklist of questions to investigate, not as a final negotiating position. The job of the AI here is to surface what to look at, not to make the call for you.
Jurisdiction matters because it decides whose courts and whose law handle a dispute. As a small business, prefer your home jurisdiction or a neutral one over the counterparty's faraway courts. If the counterparty is much larger and insists on their jurisdiction, the trade is usually fine for small contracts but is a real cost on bigger ones because litigating abroad is expensive.
Hire a real lawyer for employment agreements, equity and cap table documents, fundraising, licensing of meaningful IP, anything involving regulated industries (healthcare, finance, kids), and any contract whose worst plausible outcome could meaningfully damage your business or trigger a personal liability claim. For everything else, a clean DIY workflow with AI drafting plus a careful read is usually proportionate.
Contracts are one piece of the back office work that quietly eats a founder's week. The same pattern (draft, review, sign, store, renew) shows up in invoices, expenses, vendor onboarding, compliance reminders, and a dozen other admin chores that nobody enjoys and nobody can delegate to a cheap freelancer reliably. If contracts felt heavy this article, the rest of the back office is probably heavier. The next read is the practical companion: how to route the entire admin function through one AI Employee so contracts become one chore in a larger system instead of a recurring fire drill.
The takeaway is not that lawyers are optional, it is that most simple business contracts do not need a lawyer's hour to exist, they need a clean process, a few good templates, and someone (or something) that keeps track of versions and renewals. AI drafting collapses the cost of producing version one to almost zero, which means you can finally afford to put every small relationship on paper instead of relying on memory and goodwill. Save the lawyer budget for the deals where the worst plausible outcome is genuinely scary, and write everything else yourself with confidence. If you want to make this a system rather than a one off, an AI Employee inside Sistava can hold your templates, redline counterparty drafts, and remind you about renewal dates so contract work stops being something you dread on Sunday night.