IN Brief:
- Loft introduces an AI developer agent to turn SOPs into automated workflows.
- External intelligence is pulled from the FourKites Intelligent Network.
- Platform aims to preserve decision logic, approvals, and operating context.
FourKites has launched Loft, an AI orchestration platform designed to automate workflows across enterprise systems, supported by “Sophie,” an AI developer agent intended to convert standard operating procedures into production-ready automations. The company unveiled the platform at Manifest 2026 in Las Vegas.
FourKites is positioning Loft as an attempt to move beyond visibility layers that identify problems, but still leave teams to execute through email chains, spreadsheets, and chat threads. The company says Sophie can take operational requirements described in natural language and translate them into workflows that run across systems, with the aim of cutting deployment timelines from months to days, and reducing the ongoing engineering effort required to maintain automations over time.
Josh Jewett, operating partner at NewRoad Capital Partners and former CIO of Dollar Tree and Family Dollar, said: “Most enterprise operations still run on fragmented systems held together by spreadsheets, shared inboxes, and email chains. In that environment, critical decisions don’t live in systems at all — they live in Slack threads and people’s heads. When AI is layered on top of that fragmentation, it can observe problems but can’t reliably act on them.”
Loft’s architecture is built around three elements. Sophie is positioned as the development layer, assessing whether existing workflows can be configured, whether components can be combined, or whether custom code is needed, with FourKites engineers reviewing before deployment. “Agent Operating Procedures” are intended to record why decisions were taken and who approved them, so that reasoning can be reused when similar situations reoccur, rather than being lost in message threads. External intelligence integration pulls network-level signals into workflow decisions, such as supplier performance history and wider disruption patterns.
Mathew Elenjickal, Founder and CEO of FourKites, said: “We didn’t build AI features on top of legacy software. We built an AI-native system from the ground up. When our AI agents do the work, we record how decisions were made — not just what happened, but the context, the prior cases that informed it, and who approved it.”
FourKites is also leaning heavily on data scale as the differentiator, stating that its Intelligent Network spans more than 500,000 trading partners across 176 countries, with three million daily events across the supply chain ecosystem. The company says Loft is built to operate across ERP, ITSM, TMS, WMS, and CRM environments, and can pull data from more than 200 TMS providers, alongside other enterprise platforms.
Custom agent development through Loft is now available, with FourKites framing the platform as a foundation for a “digital workforce” that can scale repeatable decisions, rather than simply track them.
