Day 7 1. WHAT THE MARKET LEADER IS PUSHING One of the most capable legal-AI research assistants on the market now sells itself as an all-in-one "AI engineered for lawyers": legal research, litigation analysis, contract review and judicial profiling across 100-plus jurisdictions, on a free trial, pitched to everyone from solo practitioners to big law. 2. WHAT CHANGED - It's being absorbed into a practice-management giant. The product is now marketed as part of a large practice-management company that acquired it — folding a research tool into a workflow ecosystem. The likely play is distribution: bundling the AI into the parent's vast base of small and mid-sized firms rather than selling it standalone. - A new enterprise tier. A recently launched customisation offering lets large firms tailor how the AI is deployed across the organisation — a clear move up-market toward big law and in-house teams. - The last month was a market-education blitz, not a product launch. The most visible recent activity was hands-on, in-person workshops run through its European arm at bar-association libraries — teaching lawyers to use what already exists, rather than shipping something new. - Strikingly quiet in English. Over the past 30 days it barely surfaced in first-party English-language conversation. Where it did appear, it was usually name-listed alongside rival tools rather than driving its own story. 3. THE GAP The loudest conversation in legal AI right now isn't about features — it's about trust. A widely-shared finding that most users abandon an AI tool after a single mistake drew thousands of upvotes; the "AI slop" backlash is growing; and lawyers are increasingly demanding verifiable citations. The market leader sells breadth — jurisdiction count — but offered no visible answer to that verification anxiety this month, and its energy is concentrated in Europe. That is the opening. A jurisdiction-first, citation-verified legal AI — one where every answer is traceable to a real, named judgment and the underlying authority is checked — competes on exactly the axis the incumbents are ignoring: trust and local depth, not jurisdiction count. In smaller national markets the global players treat as a rounding error, provenance beats coverage.

Day 6: TASKS - [ ] Research hosting options for the app + its Postgres database — find the best way to host both - [ ] Research SA-viable payment solutions (Stripe isn't available in South Africa) — list options and any workaround - [ ] Decide whether to open the waitlist now (pre-deployment) or wait until the app is live - [ ] Work out the marketing starting point — who to make the first phone call to - [ ] Define how to test market appetite — decide a way to prove there's demand before going further DECISIONS - No firm decisions yet — hosting, payments, and marketing are all still open questions. - Accepted as fixed: Stripe can't be used from South Africa. PARKED - Incorporating abroad (Cayman Islands or Delaware) purely to unlock Stripe — idea only; don't act until the SA-payments research is done - Any real marketing push until the app is actually deployed (?) ---Two flags: I read "Venter method" as Benter method (?) and "post Grace" as Postgres — correct me if wrong.

Day 5: 1. X / tweet Summarise your legal argument on one page of A4. If it doesn't fit, leave it out. The point that wins is the one you can state plainly. The judge won't go digging for it. Length isn't rigour. Usually it's the opposite. --- 2. LinkedIn post If you can't get your legal argument onto one page of A4, you don't have your argument yet. Heads of argument run to forty pages. The case usually turns on one point. So I write the whole thing on a single page first. If a point won't fit, it goes. Not because the court is impatient — because the point I can't state simply is the point I haven't won yet. A judge doesn't reward volume. She rewards the argument she can repeat back in her own words. Find that. Lead with it. Cut the rest. --- 3. 30-second video script Most cases are won on a single point — then buried under forty pages of everything else. So try this. Write your legal argument on one page of A4. If it doesn't fit, it doesn't go in. Because the point you can't state simply is the point you haven't cracked yet. A judge doesn't reward the longest heads of argument. She rewards the one she can repeat back in her own words. Find that sentence. Lead with it. Cut the rest.

Challenge day 4: KARO surged +12.76% to $53.87 in its last session, up from a prior close of $50.60, a move well above the usual, 3% flag threshold and a clean break through the $50 level. No same-day news catalyst turned up in a search (recent headlines were about the July 28 AGM notice and last month’s Q4 FY2026 results, which already reported accelerating subscription revenue). Worth watching whether the $53-54 area holds as new support or the move partially unwinds next session. A Gmail draft (“KARO morning read 2026-07-05”) is ready in Drafts with the full readout, no buy/sell/hold framing, just what changed and what to watch.

@Matthew Lopez yes I am

@Seb Reynolds thank you for the advice