The Starting Point
OpenField Agri was a two-person founding team — one agronomist with 15 years' field experience and one commercial lead. They had a clear problem statement: UK arable farmers make critical planting and treatment decisions based on experience and gut instinct, not data. Modern precision agriculture sensors exist, but the software to make sense of the data for the average farmer is either too complex or too expensive.
They had a 20-page pitch deck, a prototype built in Notion and Figma, and 12 farmers who'd verbally agreed to try a beta. What they needed was something real — quickly enough to matter for an upcoming investor pitch.
When they came to us, they had 11 weeks until the pitch. We had 10 weeks to build something fundable.
What We Built
Week one was a joint scoping sprint. Together we cut their feature wishlist from 60+ items to 12 that formed the "core loop" — the minimum experience a farmer needed to see the product's value. Everything else went to a backlog for post-raise.
Field Mapping
Farmers draw field boundaries on an interactive map (powered by Mapbox). Each field stores crop rotation history, soil type, and area. Intuitive enough to use on a mobile phone in a tractor cab.
IoT Sensor Ingestion
Webhooks and MQTT ingestion for soil moisture, temperature, and rainfall sensors from three common hardware brands used by the beta farmers. Data normalised into a unified schema and displayed as time-series charts per field.
Yield & Treatment Tracking
Farmers log spray applications, fertiliser events, and harvest yields per field. The system calculates cost-per-tonne and flags yield anomalies by comparing against historical averages and neighbouring fields.
Alerts & Recommendations
Rule-based alerts (soil moisture below threshold, frost risk from weather API, optimal spray windows) delivered as SMS and push notifications. The agronomist founder's expertise encoded as rules the MVP could surface.
How We Worked
We ran five two-week sprints with a strict rule: no new features added mid-sprint. The founders were deeply involved in daily standups for the first three sprints, then transitioned to weekly demos as we moved into build phase.
One decision we pushed back on: the founders wanted a sophisticated AI yield prediction model in the MVP. We convinced them to replace it with simpler rule-based alerts for v1 — enough to demonstrate the concept to investors without the complexity that would have blown the timeline. The ML roadmap was documented and included in the pitch deck as "what Series A unlocks."
Technology Used
Results
"Every investor we pitched asked to see a live demo. Webmatx made that possible. Having a real product with real farmers using it changed the conversation completely — we went from 'interesting idea' to 'here's a term sheet' in three pitches."
