IICA Digital Agriculture Week 2025 — AI Day as institutional-convening anchor for LAC agrifood AI deployments
Global (IICA member states across the Americas; 2025-10 event hosted at IICA HQ San José, Costa Rica)
Content
The AI Day at Digital Agriculture Week 2025 (DAW 2025), convened by the Inter-American Institute for Cooperation on Agriculture (IICA) together with the Inter-American Development Bank (IDB), CAF — Development Bank of Latin America and the Caribbean, Bayer, PROCISUR (Cooperative Program for the Technological Development of Agrifood and Agro-Industry in the Southern Cone), the University of Córdoba, Spain, and AWS — held at IICA HQ in San José, Costa Rica, October 2025 — is the most important institutional convening for LAC agrifood AI in the corpus’s 2025-2026 window.
Documented scope (per IICA primary press release, October 2 2025):
- 200+ participants in hybrid format (in-person + virtual)
- Event name: “Innovating Agriculture with AI: From Vision to Action”
- Two-day format with nine high-level presentations
- Topics covered: precision farming, automation, data use, connectivity, productivity, sustainability, organisational culture
Named AI deployments surfaced at AI Day (cross-units in this corpus):
- PineSORT — Drone-based pineapple plant counting (presented by Fabián Fallas Moya, UCR + TEC). Substrate unit:
pinesort-ainnovatech-costa-rica-pineapple-ai.md. - AinnovaTech — Computer vision for pineapple flowering counting (presented by Rodrigo Herrera Garro, co-founder + CTO). Substrate unit:
pinesort-ainnovatech-costa-rica-pineapple-ai.md. - Autoplants (Argentine) — AI-driven rainfall/frost/irrigation prediction (presented by Julián Mauro, co-founder + Director). Substrate unit: not yet drafted; tracked as candidate for next cycle.
- Tomás de Camino Beck, Director, School of Intelligent Systems, Cenfotec University Costa Rica — opening keynote on success stories and impact.
- Fulvio Lizano Madriz (UNA + UNED Costa Rica) — Prototype smart hive for meliponiculture measuring temperature, humidity, light and air quality. Not yet agrifood-AI unit; tracked as prototype.
- Saúl Calderón Ramírez, School of Computer Engineering, TEC — on public R&D investment as prerequisite to AI-driven inclusion.
- Jesús Bejarano Quesada, UCR — “Data is the gold of artificial intelligence. Without reliable information, any model loses value.”
- Adrián Miranda Cordero, Senior Solutions Engineer, Cloudera — “Technology is not the enemy; it is an ally for making better decisions.”
Co-conveners’ roles (per IICA release):
- IICA — institutional host and convening
- IDB — finance + programming for the LAC region (independent of US)
- CAF — finance + programming for the LAC region
- Bayer — corporate-vendor seed-and-crop-protection player
- PROCISUR — technical cooperation across southern cone
- University of Córdoba (Spain) — academic-research collaboration
- AWS — cloud-infrastructure and AI-services vendor
The seven-co-convener pattern (multilateral institution + two regional banks + corporate vendor + technical-cooperation body + academic partner + cloud-vendor) is a distinct institutional pattern the corpus should surface. The IICA DAW 2025 convening is not a single-vendor-led event, nor a single-government-led event. It is a multi-stakeholder convening with the multilateral institution (IICA) as the institutional host and orchestrator.
Why a unit, not just a section in the regional scan:
- Concrete scope that the corpus can cite directly. IICA member-state geography; named co-conveners with specific roles; documented numbers (200+ participants, nine high-level presentations); specific deployment namesurfacing from the event.
- Multi-vendor and multi-institutional surfacing role. The corpus’s prior exposure to LAC AI deployments has been vendor-led surfaces (Agrosmart, Kilimo via Microsoft, Falabella via Google Cloud). IICA DAW 2025 demonstrates institutional-event-led surfacing of the same deployments. This is a structurally-distinct pattern.
- Surface for cluster pattern candidates. The cluster of deployments surfacing at AI Day 2025 — PineSORT, AinnovaTech, Autoplants, plus the Autoplants-style Argentine agtech — begins to converge on a Latin American AI-for-agriculture cluster pattern, distinct from NA / EU / China / India / Japan / Korea / Oceania patterns.
- Cluster-pattern observation value. IICA’s convening role positions Latin American agricultural AI as emerging from a regionally-coordinated institutional substrate, not a vendor-led or state-led single-deposit pattern. This is structurally distinct from NA (vendor-led) and from India (state-DPI-led) and from China (state-vendor-hybrid + provincial). The “LAC cluster pattern” candidate is one of the corpus’s working hypotheses worth tracking.
Strategic significance:
- The IICA institutional layer is the corpus’s first multi-stakeholder multilateral-convening anchor for agrifood AI. IICA’s 34 member states across the Americas makes this institution the continental-scale institutional substrate that the corpus previously had no representation for. The unit anchors the institutional substrate layer of the LAC cluster pattern.
- The pro-AI-with-conditions framing (IICA: “AI should empower rather than exclude. The region must take ownership of these technologies, developing them in an open, accessible way and adapted to its reality”) is a prospective-institutional frame, distinct from academic or civil-society critical voices. Worth surfacing in any talk about how the cluster-pattern taxonomy maps to AI governance.
- CAF + IDB + IICA represents the multilateral-finance-and-institutional layer of LAC. Distinct from the OECD-anchored multilateral development banks that anchor EU-style deployment, and distinct from China’s bilateral-and-tied-aid pattern, and distinct from US USAID-style aid. Worth a cluster-pattern layer of its own.
What this unit is doing in the taxonomy
Anchors the institutional-convening multilateral layer of the LAC cluster pattern. Distinct from:
- USDA-NIFA / NSF AI Institutes cluster (
us-academic-research.md,cornell-atkinson-idsov-cluster.md) — NA institutional substrate - WAGRI (
wagri-japan-agricultural-data-platform.md) — Japanese state-DPI substrate - AgriStack / DAM (
india-digital-agriculture-mission-agristack.md) — Indian state-DPI substrate - Korea Smart Farm Innovation Valley (
korea-smart-farm-innovation-valley-rda.md) — Korean state-anchored cluster programme - WAICO / Shanghai multilateral AI (
waico-alliance-china-multilateral-ai.md) — Chinese state-multilateral coordination reach
IICA is the multilateral-convening layer for the LAC region — a working hypothesis of an eighth cluster-pattern observation: LAC may be primarily multilateral-institutional convening in its agrifood-AI surfacing pattern.
Why it matters for talks
- Most concrete demonstration of LAC AI institutional substrate the corpus has.
- 200+ participants, 9 presentations, seven co-conveners is a strong specific-surface pattern for any talk that needs to ground the LAC institutional layer.
- The PineSORT + AinnovaTech cluster surfacing at DAW 2025 is a teachable pattern observation: institutional convenings surface cluster patterns that vendor-led searches don’t.
- The multi-stakeholder convening pattern is itself a structural observation distinct from single-vendor-led NA-style deployment convenings (e.g. AWS re:Invent) and from single-government convenings (e.g. India’s IndiaAI Mission).
Critical context
- This is an institutional anchor unit, not a deployment unit. Activity status: institutional-anchor; this is a third category alongside deployed and piloting. Worth surfacing as a deliberate schema extension.
- The “AI Day” framing emphasises adoption and acceleration. Critical voices (like IT for Change / FIAN’s Monsalve-Suárez and Seufert 2022 “Big Tech Takeover of Food Systems in Latin America”) were not surfaced as featured speakers at this event. The cluster-pattern-as-emerging-from-institutional-convening observation should be tempered with awareness that the convening itself is pro-AI-adoption in framing.
- The 200+ participants figure is institutional participation, not deployment scale. Deployment-scale figures live in the substrate units (PineSORT, AinnovaTech, Kilimo, Agrosmart, Falabella TARS).
- Latin American institutional-AI events reach a wider audience via IICA than individual vendor events. Trade-press coverage is sparser than for NA equivalents; institutional-event primary sources (IICA press releases, partner sites) are the documentation substrate.
- The cluster-pattern observation about LAC (multilateral-institutional substrate) is a working hypothesis, not a closed typology. Future cycles should test it against other LAC AI events (e.g. World Agri-Tech South America Summit 2024; bilateral-context IDB-specific programmes).
- Convening vs deployment framing caveat. The IICA DAW 2025 event was explicitly positioned as “From Vision to Action” — meaning the cycle of convening → action is forward-looking. The corpus should not over-claim deployment scale from the convening.