BergAuto · Enterprise workflow automation
The automation layer across the system.
BergAuto automates repeatable steps while approvals, exceptions, confidence checks and audit logs stay visible.
- 0.97Classify
- 0.94Extract
- 0.91Route
- 0.99Format
- Score
Every action carries a confidence score.
- Gate
Below 94, a person decides; the band sets who.
- Ledger
Every run, score and decision written to BergFlow.
- 100%Of actions carry a confidence score
- 4Escalation paths, from QA sample to specialist
- 0Unaudited auto-actions
Design figures: they describe how every BergAuto workflow is built, on every engagement.
The Trust Gate
Approval and exception rules, set before the first run.
Every BergAuto step scores its own output. The score decides the route: proceed, sample, team-lead review or specialist review. The bands below are the defaults; each workflow tunes them from shadow-run evidence, and every change is versioned.
BergFlow logs the outcome of every route, so the audit trail exists whether or not a person ever had to look.
How it runs
One master workflow, scored at every step.
01Workflow-level orchestration
One master workflow per process: intake, AI steps, conditions, waits and outputs on one canvas, run in batches or live.
02Confidence-aware governance
Each step scores its own output. The score decides the route, and the thresholds are agreed with you before go-live.
03Recorded decisions, reusable logic
Every run is written to a ledger with inputs, scores and the reviewer’s decision, so logic that works once can be reused across workflows.
Common workflow surfaces
Where BergAuto runs most often.
Four workflows from the use-case library, each with its AI step, its review path and an example metric.
- Finance
AR collections workflows
- AI step
- Prioritise and draft follow-up
- Review
- Finance approval
Example metricDSO -12% - HR
Employee onboarding workflow ownership
- AI step
- Workflow orchestration
- Review
- Document verify
Example metric40% TAT cut - CX
Escalation handling + approvals
- AI step
- Route by severity
- Review
- Manager approval
Example metricEscalations -28% - Sales
ABM campaign QA
- AI step
- Check audience and copy
- Review
- Human signoff
Example metricQA defects -54%
Capability benchmark
Built for governed work, not just faster work.
Native means built in and operated by BergLabs. Partial means possible with extra build. Limited means not designed for it.
| Capability | BergAuto | Open-source orchestration | In-house build | Model builder |
|---|---|---|---|---|
| Visual canvasOne canvas for intake, AI steps, conditions, waits and outputs | Native | Partial | Limited | Limited |
| Batch executionThousands of items run through the same workflow | Native | Native | Partial | Partial |
| Multi-format intakeEmail, documents, voice, images and structured data | Native | Partial | Partial | Partial |
| One master workflowOne governed definition per process, versioned | Native | Partial | Partial | Limited |
| Human wait statesRuns pause for a named reviewer and resume | Native | Limited | Partial | Limited |
| Confidence-gated routingEvery step scored and routed by band | Native | Limited | Partial | Limited |
| Run ledgerInputs, scores and decisions recorded per run | Native | Partial | Partial | Limited |
| Operating ownershipA named team runs the workflow to an SLA | Native | Limited | Partial | Limited |
Evidence note: BergLabs assessment of each category as typically deployed, October 2026. Individual products vary; ask us for the criteria behind any cell.
The automation boundary
The boundary is a setting, not a guess.
Repeatable steps with a known rule and a clear threshold are automated. Exceptions, regulated calls, high-value decisions and uncertain outcomes are reviewed. Where the line sits is agreed before go-live and moved only with evidence.
- Shadow run first: scores are compared with your team’s decisions
- Thresholds set from what the shadow run shows
- Every change to a threshold versioned in BergFlow
Automate · 94 and up%
Proceeds automatically and is still logged in the run ledger.
Review · 85 to 94%
A QA reviewer samples the output before it counts.
Review · 70 to 85%
A team lead checks it and records the decision.
Review · Under 70%
A specialist decides. This band is always human.
Questions
Questions about BergAuto
Scoring, thresholds and what happens when confidence is low.
Scoring and review
What does BergAuto automate?
Repeatable steps with a known rule: classification, extraction, routing, formatting, enrichment and system updates. Each step is scored, and only steps at or above the agreed threshold proceed without a person.
What happens when confidence is low?
Below 94%, the item is routed by band: 85 to 94% goes to a QA sample, 70 to 85% to a team lead, and anything under 70% to a specialist. Every route is logged in BergFlow.
Can the thresholds be changed?
Yes. Thresholds start from the Trust Gate defaults and are tuned per workflow from shadow-run evidence, with every change versioned.
Audit and systems
Is anything automated without an audit trail?
No. Every automated action is scored and written to the run ledger, so there are zero unaudited auto-actions.
Which systems can BergAuto act in?
BergAuto reads and writes through APIs, webhooks and database connections to the systems a team already runs, such as CRM, ERP, ticketing and document stores.
Start with the workflow that repeats every day.
One workflow, scored at every step, with a person on every exception. We run it in shadow first and show you the numbers.