EVAA · VENTURES
OpenRift
AI operations for teams that ship faster than they grow.
OpenRift is EVAA's AI operations venture — Slack-native automations and agentic workflows for boutique teams. Small agencies, small banks, small consultancies, and the IFA workflows that drown a two-person ops team in email threads. Less ops work, more strategic surface area.
WHAT WE AUTOMATE
The work that scales people, not the work people scale.
OpenRift focuses on the operational layer that boutique teams quietly drown in — inboxes, recurring approvals, status chasing, document handling.
Small consultancies, agencies, and banks live in two places: Slack and the inbox. The bottleneck is rarely the work itself. It is the work around the work — the email threads, the approval loops, the "did anyone reply to that?" message at eleven at night. OpenRift collapses the coordination layer without removing the human in the loop.
The canonical example is the independent financial adviser workflow. An IFA spends hours a day moving information between client emails, CRMs, and product platforms. An OpenRift workflow watches the inbox, drafts the reply, proposes the next-best action, and routes anything ambiguous back to a human in Slack. The human still decides. The work around the work shrinks.
The same shape applies elsewhere. Drafting outbound for a boutique agency. Triage and summarisation for a small consultancy's project queue. Calendar coordination across a partner network. The pattern is consistent: collapse the coordination overhead, keep the judgement where it belongs.
OUR APPROACH
Slack-first. Agent-driven. Audit-trailed.
The product lives where the work already lives. Every action an agent takes is reviewable, reversible, and logged.
Slack-first is an engineering principle before it is a product choice. Boutique teams already coordinate in Slack. Adding another dashboard adds another tab, another login, another reason for the workflow to fall out of use. The workflow shows up where the team already is, posts what it is about to do, and waits for a thumbs-up on anything irreversible.
Audit-trailed is the second principle. Every action — who triggered it, what the workflow saw, what it did, what it skipped — is logged where a human can read it. For a small bank or an advisory team, the audit trail is the product. Compliance and velocity are not in tension when the trail is clean.
The third principle is opinion. We are opinionated about which workflows AI should and should not touch. Drafting, triage, summarisation, and calendar coordination are good candidates. Final approvals, sensitive client conversations, and anything with a hard regulatory edge are not. The boundary moves with the evidence, but the default is caution.
Each workflow is a prompt, a tool set, and a memory of how the team actually does the job. The longer it runs, the sharper it gets — but the goal is never autonomy for its own sake. The goal is more time spent on the work that required a person in the first place.