What is SOMA Adaptive Memory Intelligence™ (SOMA AMI™)?
SOMA Adaptive Memory Intelligence™ by Memrail is a deterministic decision plane and memory engine for AI agents. It structures system state into Atoms (typed facts), applies triggers and policy-driven arbitration, and governs which tools, workflows, or actions may run, while returning a complete decision trace.
Who is SOMA AMI™ for?
Teams building LLM/agent workflows that must be repeatable, auditable, and safe—customer support, humanoid robots, RevOps, risk & trust, RPA/automation, and internal agent platforms.
What problems does it solve?
SOMA AMI™ addresses critical challenges in AI agent systems by providing a centralized, deterministic, auditable decision plane for agent decision-making. When agents produce flaky or non-repeatable decisions, SOMA AMI™ ensures consistent outcomes through deterministic selection algorithms with built-in tie-breaking mechanisms.
Multi-agent conflicts are resolved through sophisticated arbitration that prevents conflicting actions from executing simultaneously. The system maintains complete audit trails through first-class decision traces, eliminating visibility gaps that plague traditional agent architectures.
Policy drift—where agent behavior changes unexpectedly over time—is prevented through versioned registry and policy management, with optional request-time version pins ensuring stable, reproducible behavior across deployments.
What are ATOMs?
ATOMs: Adoption Translation and Observation Model — the canonical schema for memory facts. They standardize how agents express state, tags, and observations so downstream decisions are consistent.
ATOMs are typed facts that serve as the only inputs to SOMA AMI™.
Which Atom types exist in v1?
Are Atoms the only inputs to /ami/invoke?
Yes. Atoms are the input surface. For time‑window logic, SOMA AMI™ can also read persisted events that you ingest separately, but those are merged with any context EventAtoms during a single decision.
What is an EMU?
An EMU is an Executable Memory Unit. It has a trigger, action and policies associated with it.
How does SOMA AMI™ guarantee determinism?
SOMA AMI™ ensures every decision is perfectly repeatable by treating each evaluation as a pure function with clearly defined inputs and deterministic tie-breaking rules.
How does recent_event(...) work?
The recent_event(...) function evaluates whether specific events occurred within defined time windows, enabling temporal logic in trigger conditions.
What is the Trigger DSL?
A compact, boolean expression language that evaluates state conditions, tags, and recent events to determine when actions should be triggered.
The DSL supports logical operators (AND, OR, NOT) and functions for state evaluation, tag matching, and temporal event queries within specified time windows.
What are shadow and dry‑run modes?
Shadow and dry-run modes enable safe testing and validation of EMU configurations without affecting production systems or recording permanent state changes.
What is in the decision trace?
A comprehensive record including the registry/policy versions, evaluation context, inputs summary, every candidate with pass/fail reasons, suppression codes, scores, and the final selected results.
What are the EMU lifecycle states?
EMUs progress through a managed lifecycle that enables safe, staged rollout from development through production. Transitions are enforced by the lifecycle API; soft deletes only (you can read history).
How do registry/policy version pins work?
You can pin specific registry and policy versions at request time to ensure reproducible behavior across deployments. If pinned versions don't match the server's active versions, the request is rejected.
Which regions are supported?
us-east-1, eu-west-1, ap-southeast-1. Choose per operator in config.
What data does SOMA store and for how long?
SOMA stores events, decision traces, and operational metadata with configurable retention policies to ensure compliance while maintaining system performance and auditability.
Does SOMA AMI™ execute actions?
SOMA AMI™ returns a payload (e.g., tool_call) for selected EMUs; your host system typically executes the action and records telemetry back (events, state tags, etc.).
How do I register an EMU?
EMUs are registered through our REST API by submitting a trigger condition, policy configuration, and optional action payload.
How do I call /ami/invoke?
SOMA AMI™ accepts input facts (Atoms), context metadata, and configuration options to evaluate registered EMUs and return deterministic action recommendations.
What are the rate limits?
SOMA AMI™ implements rate limiting to ensure system stability and fair usage across all operators. Limits vary by subscription tier and usage patterns.
How does SOMA AMI™ handle ML‑derived facts?
Atoms can be tagged with their source, including ML-derived inputs. Operator policies can control how automatic EMUs behave when they depend on ML-sourced facts, enabling fine-grained control over automation confidence levels.
Does SOMA support no‑PII configurations?
Yes. Register ID‑only EMUs (no action); /ami/invoke returns the EMU id/intent/policy without payload. You can map EMU ids to actions inside your secure environment.
Glossary
- Atom / ATOM — typed fact; abbreviation of "Adoption Translation and Observation Model."
- EMU — Executable Memory Unit; a registered decision unit combining trigger, policy, and optional action.
- Trace — full, deterministic rationale for each decision.
- Decision Plane — the architectural layer that separates decision logic from application code.