Built by People Who Know the Field.
Incubated within the Qidian Venture Partners ecosystem. Our founding team combines 30+ years of combined experience in applied AI research, agricultural science, and emerging-markets operations. We didn't build Qidian VP from a conference room — we built it in the field, with farmers.
Founders
Ha Phung
Chief Executive Officer
15+ years leading agricultural technology programmes across Southeast Asia. Worked with IRRI-affiliated field programmes in Vietnam and Philippines as an agricultural technology consultant, supporting digital advisory tool validation. MSc in Agricultural Systems from Wageningen University.
Ngan Phung
Chief Technology Officer
5 years as a software engineer at a major cloud AI platform, where she worked on production ML pipelines for agri-food sector customers. Co-author of published research on transfer learning for low-resource image classification. Now applying that expertise to build Qidian VP's Gemini integration layer.
Vi Tien
Chief Operating Officer
10+ years scaling operations for tech companies in frontier markets. Previously GM of Southeast Asia at a Series B agri-fintech company, where he launched mobile-first products reaching 80,000 smallholder farmers across Indonesia and Vietnam. MBA from INSEAD (Singapore campus).
Technical Inspiration & Open Research
Qidian VP is built on foundational research published by leading AI labs. The following papers and projects form the scientific basis of our technology.
"Attention Is All You Need"
Vaswani et al., 2017 — The Transformer architecture paper that underpins all modern large language models, including Gemini. Qidian VP leverages Transformer-based multimodal reasoning for crop disease diagnosis.
NeurIPS 2017 · Google Research"Gemini: A Family of Highly Capable Multimodal Models"
Google DeepMind, 2024 — The model powering Qidian VP. Gemini's native multimodal capability (image + voice + text in a single call) is what enables real-time crop diagnosis from smartphone photos.
Google DeepMind Technical Report, 2024AlphaFold & Protein Structure Prediction
Google DeepMind, 2020–2024 — Proof that AI can solve fundamental life-science challenges at scale. AlphaFold's success inspired our conviction that AI can similarly transform agricultural diagnostics.
Nature, 2021 · Nobel Prize in Chemistry, 2024Google's Responsible AI Practices & Model Cards
We follow Google's responsible AI framework, including model cards for transparency, fairness evaluation across demographic groups, and Search Grounding to reduce hallucination risk in safety-critical agricultural advice.
Google AI Principles · Model Cards for Model ReportingAdvisory Board
Industry veterans and domain experts who provide strategic guidance, open doors, and challenge our thinking.
Dr. Trần Đại Nghĩa
Agricultural Science Advisor
Formerly at Cần Thơ University Plant Protection Department. Former Deputy Director, Vietnam National Institute of Agricultural Sciences (VNIAS). 25+ years leading crop protection research programmes. PhD in Plant Pathology from Wageningen University.
Sarah Chen
Cloud & AI Strategy Advisor
Based in Singapore, previously at Google Cloud for 5 years as Customer Engineer (SEA region). Now advising 4 AI startups. Specialises in Vertex AI deployment patterns and cost optimisation for emerging-market products.
Nguyen Minh Duc
Agri-Finance Advisor
CEO of TinVay Microfinance, serving rural communities in 12 provinces across Vietnam and Cambodia. 200,000+ smallholder farmer clients. Expert in financial product design for rural populations.
Dr. Somchai Rattanapong
SEA Market Expansion Advisor
Former Country Director, Swisscontact Thailand. Deep network across ASEAN agricultural ministries and cooperative federations in Thailand and Myanmar.
Research Partnerships
We collaborate with leading agricultural research institutions to ensure our AI models are grounded in peer-reviewed science and validated field data.
Nông Lâm University
Ho Chi Minh City, Vietnam
Joint research programme on tropical crop disease identification using computer vision. Providing 15,000+ annotated field images from their experimental farms and access to faculty agronomists for model validation.
Kasetsart University
Bangkok, Thailand
Collaboration on Thai language NLP models for agricultural terminology and regional dialect recognition. Co-developing training datasets for cassava and sugarcane diseases endemic to Thailand's central and northeastern regions.
IRRI — Open-Access Datasets
Los Baños, Philippines
Trained on publicly available datasets from IRRI Rice Knowledge Bank and PlantVillage open-source repositories — the world's most comprehensive open databases of rice diseases, pest identification, and treatment protocols. Used as the primary ground-truth dataset for our diagnostic engine.
Where We Are
Key milestones in Qidian VP's development — from idea to pilot deployment.
Idea & Validation
Founded in Ho Chi Minh City. Validated concept with farmer interviews in Mekong Delta — confirmed strong demand for accessible, dialect-aware agronomic advice.
Technical Prototype
Built first working prototype using Gemini 2.0 Flash. Integrated open-access crop disease datasets from PlantVillage and IRRI Rice Knowledge Bank — 120,000+ annotated images.
Pilot Launch
Deployed to 312 farmers across 3 cooperatives in Cần Thơ and Đồng Tháp. Achieved 87.3% diagnostic accuracy validated by independent agronomists using IRRI protocols.
Scaling & Fundraising
Expanding pilot to 500 farms across Mekong Delta. Building cooperative partnerships and raising a pre-seed round to accelerate growth.
Join the Mission.
We are building a small, high-conviction team of AI engineers, agronomists, and emerging-markets operators. If you believe AI should be deployed where it has the highest human impact — not just the highest margin — we want to talk.