Services

Four things, all for people who work in or sell to UK agriculture. Most engagements start with a conversation about what you are trying to do, not a proposal.

AI consultancy: putting AI to work in an agricultural business

For farming businesses, agricultural suppliers, merchants, media companies and agritech firms that want to adopt AI properly rather than buy whatever is being sold this month. We do not build software or AI agents ourselves; we help you decide what you need, choose it, and get it working in your business.

  • Finding the right starting points: where AI saves real time or opens revenue in your operation, and where it should be left alone.
  • Choosing between off-the-shelf tools, specialist agricultural products and custom work, and introducing the right partner when you need something built.
  • Implementation: rolling AI tools out to a team, setting the rules for how they are used, training staff, and measuring whether they are paying their way.
  • Data readiness and rights: getting your documents, product data and records into a state an AI tool can use, and making sure you are allowed to use them.
  • Evaluating AI models for a specific job. In our own project work we run the same question sets across Anthropic, Google and OpenAI models and compare quality, cost and speed.

Private AI on your own hardware

Some businesses cannot or will not send their data to a public AI service: commercially sensitive pricing, customer records, trial results, anything under a licence that forbids it. For them we advise on running open models privately, on hardware you own or in a dedicated UK facility.

  • Working out whether private hosting is justified for your use, and what it will cost against the public alternatives.
  • Specifying the set-up: which open models fit the job, what hardware they need, on-premises versus co-location versus dedicated hosting.
  • Hardware and hosting through our partner Danucore, who supply on-premises AI systems, co-location and UK-based AI hosting, and model selection and tuning support.

Day-rate or fixed-scope, with a short paid discovery as the usual first step.

Speaking and training

Talks, workshops and in-house training on practical AI for farming audiences. Chris has delivered AI training courses nationally and speaks to farmer groups, agricultural companies and industry events.

  • Keynotes and conference sessions: what AI can and cannot do for a farm business today, with live demonstrations.
  • Half-day and full-day workshops for teams: using AI assistants safely for research, writing, analysis and customer service.
  • Board and leadership briefings for agricultural businesses planning AI investment.

Sessions are tailored to the audience and avoid jargon. Enquire about dates and fees.

WhatsApp marketing and automation

Farmers run their businesses on WhatsApp. Cosecha Digital is a reseller for VOXD.ai, a WhatsApp business messaging platform, and helps agricultural companies put it to work.

  • WhatsApp broadcast and campaign set-up for product launches, events, price updates and seasonal advice.
  • Two-way customer conversations handled by a mix of AI and your own team, with hand-off rules you control.
  • Implementation and the integration layer: joining VOXD to your product data, CRM, order system or knowledge base so replies are accurate and current.
  • Opt-in, consent and data handling set up properly from the start.

AI products for agriculture

Alongside client work, Cosecha develops its own products. The first is a UK farming knowledge layer: one organised, rights-cleared store of trials, variety data, product registers, guidance and transcribed video and podcast content that any AI model can search, with the source recorded for every item.

It is being built and tested in the open through the AI for Agronomy trial with real farmers through a full cropping year. The design rules are rights first, never overwrite, always cite, and model-agnostic.

If you hold agricultural data or content and want it to be findable by AI tools on your terms, with attribution, we would like to talk.