EveryPig helps pork producers, farm managers, and veterinarians improve animal health and operational visibility. Its platform manages data for millions of pigs and replaces paper barn sheets with digital production records.

But collecting data and making it easy to use across multiple customer organizations are different challenges. Answering an analytical question could still require a support request, a manual export, or a developer-written SQL query. Tech 42 built an AI data agent on AWS that lets authorized users ask questions in plain English and receive answers inside EveryPig in real time.

Results: 

1. Most user questions answered by the agent without involving support team
2. Report requests that once took up to a week are now answered in seconds, inside the platform
3. No CSV exports or developer requests needed to answer analytical questions
4. Production-ready, multi-tenant platform that can scale across the customer base
5. Tenant-scoped access controls restrict queries to approved, read-only reporting views

“Before this, if one of our customers needed a new data report, our engineering team had to build it from scratch. Now, our producers, veterinarians, farm teams, and other stakeholders can just ask questions in plain language and get direct, self-service access to their data. They can investigate trends, catch emerging concerns, and know exactly where to focus — without waiting on us.

“The impact has been clear: PigPulse returns detailed findings in under a minute, has cut custom-report requests , and resolves most of our users' questions without ever coming to our team. That's freed up our engineering capacity for higher-priority work, and it means our customers are making faster, more informed decisions.”

- Chris Bomgaars, Founder, EveryPig

Challenge: Production data locked behind SQL and support tickets

EveryPig's platform holds mortality records, treatment histories, compliance status, and site comparisons for hundreds of farms. Getting an answer from that data required SQL skills, or a request to someone who had them.

Key challenges included:

• Manual data pulls: Answering a question meant a CSV export, a spreadsheet, or a request to someone who could write a query. That delayed data access for farm managers, veterinarians, and production teams.
• Data spread across tables: Mortality trends, treatment summaries, compliance status, and cross-site comparisons existed in the system, but pulling them together meant bridging multiple database tables.
• Multi-organization access: In order to be useful in production, the platform needed to have a multi-tenant architecture that respected access controls and restricted queries to approved, read-only reporting views
• Contextual insights: A generic model analysis wasn’t enough. The agent needed deep industry context built in, so the insights and language it used matched what farm managers and veterinarians actually trust.
• Time delays: Report requests common to pork production often queue for months under legacy systems. EveryPig had already moved past paper records, but wanted insights available in real time, not on a backlog of requests.

Solution: An AgentCore AI agent deployed on AWS 

Tech 42 deployed its AI Agent Starter Pack, built on Amazon Bedrock AgentCore, directly into EveryPig's AWS account. The Starter Pack is built on Infrastructure-as-Code principles, so deployment was simple and no infrastructure had to be built from scratch.

User experience was a top priority for EveryPig, so the agent is configured to handle natural language questions via chat or voice and deliver actionable insights from their organization’s data, accessed via a custom-built MCP server.  Tech 42 also customized the agent to understand critical terminology and the types of questions farm teams actually ask. 

The agent was also production-hardened with enterprise-grade security and guardrails, multi-tenant data isolation and access controls, and robust tracing and observability. And to give the EveryPig deeper insights into customer adoption, they’re about to use per-tenant usage tracking and spend controls.

Outcome: technical data analysis made accessible to farmers and veterinarians

EveryPig launched their new agent to a beta group of four of their largest customers, putting the agent in front of production-scale users.

Report requests that once sat in a queue for up to a week now get answered directly inside the platform within seconds. Mortality trends, treatment summaries, compliance status, and cross-site comparisons, all technically present in the system before, are now reachable without a CSV export or a developer. 

By delivering support for multiple customer organizations through this shared agent, the EverPig team is poised to scale across their customer base. And for EveryPig’s engineering team, this means they are freed up for higher priority work while delivering faster insights to their customers. 

Teams needing to put data analysis directly in the hands of non-technical users or looking to scale proof of concepts to multi-tenant infrastructure can review Tech 42's AI agent services or begin with an executive workshop to explore what’s possible for their business.

About EveryPig

EveryPig is a management platform for pig owners, producers, and veterinarians. Built by people who understand pork production firsthand, it helps teams catch health issues in real time instead of after the fact. EveryPig is used by some of the nation's top pork producers and manages data for millions of pigs. The platform works on any device, from smartphones to desktop computers, and integrates with the legacy systems producers already run.