Responsible AI for Commercial Lending Software
Suntell builds AI into Square 1 Credit Suite with a specific set of limits. This page explains what those limits are and why they exist.
AI designed with purpose
We do not add AI for novelty or marketing appeal. We deploy it where it removes repetitive, low-value work and gives credit professionals more time for analysis, judgment, and relationships.
That standard means we say no to a lot of ideas. If AI would not measurably reduce the work, we do not ship it.
What Ivy will not do
These are deliberate limits, built into how Ivy operates.
Ivy does not make credit decisions
Ivy does not approve, decline, price, or assess creditworthiness or repayment capacity. Asked whether a deal should be approved, she declines and redirects to what she can help with.
Ivy does not evaluate who should get credit
Protected class characteristics are not part of what Ivy receives or what she is asked to evaluate. She does not rank, score, or screen borrowers, and she is not used to decide who is offered credit or on what terms.
Ivy does not build prospect lists
Ivy does not identify, generate, or target prospects. She comments on the opportunities your officers have already entered. Deciding who to call is a decision your lenders make, and Ivy is not part of making it.
Ivy does not characterize people
In the Business Development module, Ivy describes activity and pipeline evidence. She reports what the record shows. She does not describe anyone’s attitude, effort, motivation, or ability. Managers draw conclusions about their people. Ivy gives them evidence.
Ivy does not present estimates as determinations
Probabilities, weighted values, and forecasts reflect what the officer entered and what the stage implies. Ivy says so when she uses them.
Ivy does not fill in what she does not know
When the record does not support an answer, Ivy says so rather than filling the gap with an assumption or an industry average.
Human oversight is not optional
Every AI-assisted feature in Square 1 Credit Suite puts a person between the output and the outcome. Ivy prepares, organizes, and suggests. A member of your team reviews it. Final decisions remain with the institution.
Built for institutional governance
Our goal is practical efficiency that aligns with institutional governance, data protection standards, and sound risk management. Community banks and credit unions operate under examination. AI that cannot survive that scrutiny is not useful, however impressive the demo.
Common questions
Where does AI processing happen?
In Suntell’s secure Azure environment, hosted in the United States.
What data is sent for processing?
Only the data relevant to the task. Tax identification numbers, account numbers, and customer name information are not transmitted.
Does Suntell use my institution’s data to train AI models?
No. Customer data is not used to train models.
Is my data combined with other institutions’ data?
No. Information is not pooled across institutions.
Is Suntell SOC 2 certified?
Yes. Suntell maintains SOC 2 Type 1.
Is a person involved in reviewing AI output?
Yes, on every AI-assisted feature. Ivy prepares, organizes, and suggests. A member of your team reviews before anything is final.
More from AI HQ: AI HQ · Business Development
See responsible AI in practice
See how Suntell keeps a person between every AI-assisted output and the outcome, with a live, personalized demo.