Steps & deliverables
How to start with Alltera, what you get at each step, when, and how the contract works.
Map & Decide
We map one process end to end and deliver a documented recommendation: go or no-go.
Mapping (Workflow Graph)
- Delivery
- End-to-end map, documented business case, decision and approval matrix, proposed initial autonomy level, go/no-go recommendation and deployment roadmap.
- Timeline
- 7–10 business days from access to data and systems
- Exit criterion
- A documented decision, including a no-go.
- How it is contracted
- Fixed scope.
Build & Integrate
We implement the solution across the systems you already use and take it through approval for production.
Implementation to approval for production
- Delivery
- Automations and agents integrated with your ERP and systems of record, rules you approve, a configured approval matrix, shadow mode with measurement, runbook and handover.
- Timeline
- 15–30 days to approval for production, from access and security approval
- Exit criterion
- Approval for production: functional, quality, security and data, governance and operations.
- How it is contracted
- Fixed scope, by milestones.
Run & Improve
We keep the AI layer running in production under service levels and evolve it based on what real cases show.
Support plan
- Delivery
- Observability, incident response, evidence-based autonomy changes and a monthly AI-layer report (quality, exceptions, cost, latency and evolution). The business indicator is attached for information.
- Timeline
- Continuous
- How it is contracted
- Tiered plan, with contracted service levels.
Optional and recommended: without support, the AI layer has no owner after go-live.
Govern & Scale
We standardize AI governance and extend what works to new use cases.
Governance and new use cases
- Delivery
- AI usage policies and controls, approval limits and autonomy levels standardized across processes, and new use cases from the accelerator library.
- Timeline
- New use cases: shorter than the first
- Exit criterion
- The same approval for production.
- How it is contracted
- Once a solution is already in production.
The process outcome belongs to the client.
Alltera is accountable for the delivery, quality, security and evolution of the AI layer, with service levels on what we control. The business indicator is co-measured and reported for information.
Support service levels
- AI layer availability
- Minimum accuracy defined in Mapping
- Incident response time
- Cost per transaction within an agreed range
Values set in the contract, by tier.
Contracting in Brazil
- Monthly fee in Brazilian reais
- Annual adjustment by index (IPCA or IGP-M)
- A clause for model costs, which are priced in US dollars
- Timelines count from access to data and systems, not from signature