Sidetrade has spent fifteen years accumulating one of the largest Order-to-Cash datasets in the world. The next step is to turn that data into intelligence that works alongside finance teams: agents that chase, reconcile, brief and escalate, grounded in real O2C context and running on infrastructure we control.
This post covers the platform we are building to do that, the framework underneath it, and the two deployments of it that exist today: Aimie IQ for our customers and Sidetrade IQ for our own teams.
01 / the vision
Pillar #1 of O2C Intelligence 2030: the O2C Intelligence Platform
Our R&D vision to 2030 starts with one platform pillar, built on three commitments.
Three new AI products, one SAFE framework
Aimie IQ, the AI engagement and orchestration layer
A library of ready-to-deploy Aimie Agents
An Agent Builder Studio to build your own agents
Sovereignty
Deliver our own GPU compute
Run open-weight LLMs, fine-tuned for O2C use cases
Keep all data and intelligence private in our infrastructure
Deliver intelligence at a competitive €/Mtok
Build real O2C intelligence
Maximise our Data Lake advantage
Self-learning: community-built skills, memory and knowledge
An Intelligence Hub for vendor-specific intel such as scorecards
Launch an O2C Eval
02 / the product family
Three products on one framework
Aimie Agents are ready to deploy for the repeatable work of collections — the first is the Aimie Cash Collection Agent. Customer Agents are the ones you build yourself in Agent Builder Studio, composed from skills and policies and published to the Agent Registry. Aimie IQ is the engagement and orchestration layer above both: one conversation that reaches skills, agents and actions.
The three are deliberately not separate products with separate runtimes. SAFE is the common foundation under all of them, which is what lets a skill built for one show up in another.
03 / the enterprise foundation
SAFE: Agent Connect and the Agent Harness
Agent Connect is how work reaches an agent: the Agent Registry as the configuration model, an event router, the HTTP surface behind the Aimie IQ chat, and the Skills Registry of O2C skills.
The Agent Harness is what happens once it does: context assembled from O2C data, capabilities exposed as skills and sub-agents, reasoning on our on-premise LLM, tool selection over O2C actions, policy management with deterministic escalation, and audit and memory so the next run is better. Pydantic AI and Temporal sit underneath.
04 / drinking our own champagne
Two IQs. A race to scale intelligence
The same platform runs in two places. Aimie IQ is the production deployment our customers use every day. Sidetrade IQ is the internal deployment: our teams and business power users build skills and agents, and use the chat for their own work.
Every skill we build for ourselves is one step closer to a skill for our customers.
SAFESidetrade Agentic Framework for Enterprise. The common foundation under all three.
Rapid scale
One declarative configuration approach. YAML defaults and custom agents through Agent Builder Studio.
Designed for enterprise
A custom framework on our on-premise LLM, with audit capability across agent execution.
Privacy and security first
Policy controls, guardrails and deterministic escalation. Every agent is validated against the OWASP Agentic Top 10.
05 / what it means day to day
Save hours every day
01
Hours back, every day
Sidetrade IQ takes the repeatable work so people can focus on judgement.
02
Agents built for the role
Agents for each role and workflow, learning from how they are used.
03
Connected to the tools people already use
Salesforce, Marketo, Oracle and the product knowledge base are reached through connectors.
06 / a system that learns
What scales what
The parts of O2C Intelligence do not grow on the same clock, and that is the point. Skills and memory scale with the user base. Agents scale with the product and the customers running it. Knowledge and the Data Lake scale as vendors are onboarded. Fine-tuned models scale with R&D.
Because all of them sit on one enterprise framework, growth in any one of them compounds into the others rather than forking into a second system to maintain.
