Central Asia B2G Intelligence
This open research catalogues 100 deployable AI and digital-government initiatives across Uzbekistan and Kyrgyzstan, scored on five axes and grounded in 100 decrees, 105 institutions, 117 decision-makers, 49 donor programmes, and 100 global precedents. Every claim is typed, verifiable, and cross-referenced. 28 initiatives are rated Tier-A — deal-ready.
Where should you start?
Pick the entry that matches what you do — each card lands you on the page that answers your first question.
Vendor / B2G operator
Find Tier-A initiatives with verified buyer, decree anchor, donor co-financing, and a credible 12-month deal path.
Donor / IFI counterpart
49 active and pipeline programmes from World Bank, ADB, EU, UN agencies, and bilaterals — TTL or PM named on each.
Investor / VC
200 solopreneur MVPs and 28 Tier-A B2G initiatives, scored on five axes with local market fit and Russian/CIS substitution lenses.
Government / regulator
How peers are deploying AI in courts, tax, health, and digital identity — with tournament-ranked transferability scores.
Researcher / journalist
Every prompt, every source URL, every audit finding is public. Reproduce the whole pipeline yourself for under USD 20 of paid API calls.
What is in this research?
A typed, source-cited knowledge graph of AI and digital government opportunities in Uzbekistan and Kyrgyzstan, plus the methodology used to build it. Each initiative is mapped to a specific decree, institution, decision-maker, donor programme, and global precedent. Numeric claims carry a verified URL; uncertain claims are tagged L2_VERIFIED or INFERRED rather than dropped.
How was it produced?
An eleven-agent multi-wave pipeline built on Claude (Opus + Sonnet) with cross-model verification via Perplexity Sonar Deep Research on a strict USD 20 OpenRouter budget. The seven waves move from blueprint to legal corpus to institutions, donors, procurement, trends, people, synthesis, adversarial audit, and outreach. The audit wave caught four wrong Tier-1 identities before publication; corrections are public.
Who built this?
Alexandr Valuev — independent researcher focused on B2G market intelligence in Central Asia. Reachable at valuev.alexandr@gmail.com · LinkedIn · GitHub. The full prompt set, agent definitions, and source code are openly licensed under Apache 2.0.