Questions about how Alltera operates.
Direct answers on what you buy, how long it takes, and how to start.
What does Alltera deliver?
We decide, build, run and govern the AI layer of your critical processes: Map & Decide, Build & Integrate, Run & Improve and Govern & Scale. See the solutions →
Who is accountable for the process outcome?
The client. Alltera is accountable for the delivery, quality, security and evolution of the AI layer, under contracted service levels.
Do you sell software?
We do not sell off-the-shelf software. Alltera Core is middleware tied to the project: it is deployed inside it and stays with you afterwards, under a continuity contract. We use accelerators built on it to put your process into production faster.
Is Alltera a BPO with AI?
No. We do not run the client's process or staff seats. We build and support the AI layer in your systems. First-level support routine (L1) stays with you or with whoever already handles it.
What can AI decide on its own?
Only what the autonomy levels allow, by decision type: from L0 (shadow, no effect) to L3 (autonomous by exception). The level rises with evidence and steps back if quality drops. Above that, the delegation of authority you approve applies. See the Control Model →
How long to production?
Mapping: 7–10 business days. Build & Integrate: 15–30 days to approval for production. These are typical ranges, counted from access to data and systems and security approval, not from signature.
How does engagement work?
Three steps: Mapping (fixed scope), Solution in production (Build & Integrate by milestones and, optionally, tiered Run & Improve with service levels) and New use cases. The initial 30-minute call is not Mapping. See steps & deliverables →
How do you handle data and LGPD?
Your data and systems of record stay with you. We process data as an operator under LGPD. Every automated decision has an approval limit, an audit trail and an owner. See Trust and security →
What is approval for production?
The five criteria a solution meets, with evidence, before it goes live: functional, quality, security and data, governance and operations. The outcome is approved, approved with restrictions, or not approved. Details in Method →
What is an FDE?
FDEs (Forward Deployed Engineers) are the engineers who work inside the client's operation. They connect systems of record and put AI automation into production. They do not replace the department or staff seats.
Terms that appear in this operation.
Case
The unit of work: what enters the process, gets decided and leaves on record.
AI layer
The automations, agents, rules and controls Alltera builds and supports on top of your systems.
Control Model
The approval for production, autonomy levels and the delegation of authority.
Approval for production
The five criteria a solution meets, with evidence, before it goes live: functional, quality, security and data, governance and operations.
Shadow mode (L0)
AI runs in parallel with no effect, and we measure agreement with whoever decides today.
Autonomy levels
L0 to L3 by decision type. They rise with evidence and step back if quality drops.
Delegation of authority
A decision matrix by value and risk, approved by the client.
Baseline
A measure of the current state (throughput, quality, cost) before any build.