Startup Base
Events
News
•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode•System is currently running in test mode
Startups
Startup Base

Platform helping founders build, validate, and scale successful startups.

+998 99 353 11 99info@startupbase.uz
4, Tepamasjid str., Tashkent, Uzbekistan, 100164

Ecosystem

  • Startups
  • Investors
  • Mentors
  • Partners
  • Accelerators
  • Incubators

Programs

  • Digital Startups Program
  • Local and International Programs
  • Register of Successful Startups
  • Crowdsourcing Platform

Events

  • Events

News

  • News

Get Support

  • Startup Advice

Resources

  • Community

Community

  • Testing your MVP

© 2026 StartupBase. All rights reserved.

Dalil AI

Dalil AI

Dalil is a case-preparation engine for advocates in Uzbekistan. It turns the facts of a case into the matters to be proved, the evidentiary gaps and a draft claim. Every citation opens the exact article on lex.uz and is verified word for word against the law, so it cannot be fabricated.

Information not available!

B2B
Pre SeedLegalTech
Tashkent, UzbekistanFounded 09.08.2026Pre Seed • —Website
0
Users
$0
MRR $

Financial Overview

$0
Valuation
$0
Enterprise Value
$0
Revenue
$0
Last Round
N/A
Investment Date

About the startup

An advocate loses a case in preparation, not in court: what must be proved is scattered across codes and Plenum rulings. General AI invents article numbers, and on lex.uz the Russian text lags behind the authoritative Uzbek one. Dalil is deterministic — the analysis is computed by a matter-to-be-proved matrix, not a model. It allocates the burden of proof, lists gaps, drafts claims and motions, flags stale translations, and refuses to start if a citation differs from the law by one word.

Problem

An advocate loses a case during preparation: what must be proven is scattered across codes and Plenum resolutions and is assembled from memory. General AI invents article numbers, and on lex.uz, the Russian text of the law lags behind the Uzbek text read by the court.

Solution

Dalil transforms the fable into a subject of proof with shared burdens, gaps, questions to the creditor, and a draft lawsuit. The analysis is determined—it is calculated by a matrix, not a model—and every citation is verified word for word with the law, which is why it cannot be fabricated.

Business model

Bar associations and law firms in Uzbekistan — B2B, priced per advocate seat, entered through a free pilot. They do not pay for access to the law, free on lex.uz, but for preparation hours and for the mistakes that cost a case: a missed element of proof, a wrong fee, a missed deadline.

Target Market

Bar associations and law firms of Uzbekistan B2B the tariff depends on the location of the lawyer's work, admission is free through a pilot. They pay not for the law, but for preparatory hours and mistakes that result in a loss of work: a missed item, an erroneous duty, a missed deadline.

Why We Need Investment

Information not available!

Meet the Team

Similar Startups in LegalTech

Other startups in the same industry you might be interested in.

Contact Details

Website
Tashkent, Uzbekistan

Achievements

Built from zero since 9 August 2026 by a team of three, and already working end to end. We assembled our own corpus of Uzbek law without an API, because lex.uz has none: 227,369 fragments from 2,017 acts — 1,172 Russian and 845 Uzbek documents, 25 codes and 103 Supreme Court Plenum rulings. Nine case-category matrices produce a deterministic analysis. Citation verification runs at every start-up and stops the engine if a citation differs from the act by a single word
it caught a fabricated citation in our own first matrix. The system detects where the Russian translation lags behind the authoritative Uzbek text, and returns an article as it stood on the date of the events. Our parser found and fixed a defect that silently dropped half of the Administrative Offences Code: articles inserted by amendments carry superscript numbers. 22 screens in production, 18 test suites, search in 10–120 ms, full case analysis in about 3 seconds. The code is public on GitHub.

Startup presentation

Download the presentation to learn more about the startup's vision, business model, and growth strategy.

Download presentation
3
Employees
$0
Investment Needed
0
Monthly Visits
KU
Kucharov Kuvonchbekvonc
QA Lawyer
AB
Abdimuratov Baymurat
Engineer
Safelaw
Safelaw

SafeLaw is an AI platform for fast and convenient access to reliable legal information. It operates on RAG technology based on the lex.uz database: 20 codes (7,056 articles), 861 laws, resolutions, and plenums of the Supreme Court. Each response is provided with a link to the original source. The platform supports 3 languages and voice input. There are tariffs for citizens, lawyers and companies, roles for B2B, positions, general balance sheet and audit log.

Pre-seedLegalTech
Dictum AI
Dictum AI

An’anaviy sud majlislari yoki yuridik maslahatlarda nutqni bayonnoma (protokol) holatiga keltirish ko‘p vaqt, kuch va inson resursini talab qiladi. Shuningdek, sohaviy atamalarning ko‘pligi sababli oddiy Speech-to-Text modellari yuridik nutqni aniq taniy olmaydi.

Pre-seedLegalTech