AI Expertise. Human Values.

Why This Page Exists

Hiring software has a trust problem. Not because teams don’t care, but because too many systems ask for faith instead of offering clarity.

This page exists to remove the mystery.

SageScreen is used in moments that matter to real people: candidates showing up honestly, and teams making decisions that affect careers, livelihoods, and culture. In those moments, “trust us” is not enough. You deserve to know how the system behaves, what it does with data, and where its limits are.

So this page is written in plain English. No marketing gloss. No abstract promises. No black boxes dressed up as magic. Just how SageScreen actually works, why it was designed this way, and where humans stay firmly in control.

This page is for customers evaluating us, candidates experiencing us, legal and procurement teams reviewing us, and frankly, for ourselves. If we can’t explain what we’re doing clearly, we shouldn’t be doing it at all.

Trust doesn’t come from perfection. It comes from transparency, restraint, and the willingness to say “here’s exactly what this system does, and here’s what it never will.”

That’s what follows.

Our Guiding Principle

Hiring decisions should be explainable, auditable, and made by people.

That principle is not marketing language. It’s a design constraint.

SageScreen exists to reduce noise in hiring, not to replace human judgment. We use AI to structure conversations, apply consistent evaluation criteria, and surface signal where interviews usually generate ambiguity. What happens next is always a human decision.

AI is good at patterning. Humans are responsible for judgment, context, and consequence. We keep that boundary explicit.

In SageScreen, the systems that conduct interviews are intentionally separated from the systems that evaluate them. And those systems are separate again from the people who ultimately decide what to do with the results. No single model interviews, evaluates, and decides. That separation is deliberate. It prevents momentum from turning into mandate.

We do not believe in opaque scoring engines or silent automation in hiring. If a result cannot be explained in plain language, it should not influence a real person’s career.

AI should calm the hiring process, not accelerate it past accountability. When a system is doing its job well, it makes decisions clearer, slower where needed, and easier to justify to another human sitting across the table.

That is the line we do not cross.

What SageScreen Is (And Is Not)

SageScreen is a structured interviewing and evaluation system designed to bring consistency and clarity to hiring conversations.

What We Do

SageScreen conducts guided, role-specific interviews using AI interviewers (“Sages”) that adapt to the conversation while staying anchored to a defined rubric. Each interview is shaped by customer-provided guides, role criteria, and guardrails, but the exact questions are generated in the moment to follow the candidate’s responses naturally.

Evaluation is performed by a separate AI evaluator that did not participate in the interview and has no influence over how questions were asked. That evaluator applies the same rubric to every candidate for the role and produces a structured result based on that single interview.

The outcome is not a hiring decision. SageScreen produces an assessment of whether a candidate Meets Expectations or Does Not Meet Expectations for the role, along with supporting context. What to do with that information is always up to the hiring team.

SageScreen also performs basic integrity checks to protect the process itself. These include presence verification, audio continuity checks, and behavioral interaction signals intended to detect obvious misuse or automation. These checks exist to maintain fairness, not to profile or judge candidates.

What We Explicitly Do Not Do

• SageScreen does not perform facial analysis. We do not measure facial features, infer demographics, assess expressions, or draw conclusions from a person’s appearance.
• We do not perform emotion detection. We do not attempt to infer feelings, personality traits, or psychological state from voice, face, or behavior.
• We do not score or evaluate appearance, environment, or presentation. A candidate’s surroundings, clothing, or physical characteristics are irrelevant to our evaluations.
• We do not make automated hiring decisions. SageScreen does not approve, reject, shortlist, or rank candidates. It provides structured insight from a single interview and nothing more.

In short, SageScreen is an evaluation aid, not an authority. It is designed to inform human judgment, not replace it.

Human Oversight by Design

SageScreen is intentionally built so no single system holds all the power.

The AI that conducts an interview is not the AI that evaluates it. Interviewers, rubric generators, and evaluators are separate agents with distinct roles and constraints. The interviewer focuses on guiding the conversation. The evaluator focuses on applying the rubric. Neither one decides what happens next.

Evaluators work from interview transcripts, not live interactions. They do not see video, hear tone, or react to the dynamics of the conversation itself. This separation reduces bias introduced by performance style and keeps the evaluation anchored to what was actually said.

SageScreen does not enforce outcomes. Customers are not required to accept, act on, or even agree with the results we provide. The most directive artifact we produce is a report or PDF summarizing the evaluation. There are no automatic gates, workflow locks, or hidden enforcement mechanisms.

Human teams remain fully responsible for decisions, follow-ups, and outcomes. That responsibility is not something we can automate away, and we do not try to.

Oversight, in this system, is not an afterthought. It is the structure.

Data Minimization, On Purpose

We collect less data because it is safer, fairer, and easier to defend.

SageScreen only collects what is necessary to conduct an interview and return results to the hiring team. For candidates, that means basic identification and contact information: first name, last name, email address, and phone number. Nothing more is required to participate.

We do not store audio or video recordings of interviews. Audio is processed in very short segments for continuity verification, compared against a baseline, and then discarded within seconds. Those segments are never retained as recordings and are not used for evaluation.

We do store transcripts and structured evaluation artifacts, because those are what make results explainable and reviewable by humans. We also retain system prompts, evaluator outputs, and intermediate artifacts that are required to reconstruct how an evaluation was produced. This is intentional. If a result matters, it must be traceable.

Customers do not configure data collection levels today. Instead, we enforce minimization by default. The system is designed so that unnecessary data is never collected in the first place.

Less data reduces risk. It also sharpens focus. What remains is signal that can be explained, audited, and challenged when needed.

Images, Identity, and Integrity Checks

SageScreen may capture a single image during an interview session. This exists for one reason: to confirm continuity and presence.

That image is used to verify that a real person is present and remains present throughout the session. In limited cases, additional images may be captured if the system detects repeated camera failures or significant continuity issues. These checks are designed to protect the integrity of the interview, not to evaluate the candidate.

Images are never analyzed beyond basic presence verification. We do not examine facial features. We do not infer identity, demographics, emotion, or intent. An image is treated as exactly that: an image confirming a person is there.

All images are retained for a limited period and then automatically deleted in batch processes. The default retention window is 30 days. Images are not used for training, profiling, or evaluation, and they are not retained longer than necessary.

Integrity checks exist to keep the process fair for everyone participating. They are guardrails, not judgments.

Bias, Fairness, and Evaluation Discipline

Bias is not eliminated by pretending systems are neutral. It is reduced through structure, consistency, and restraint.

SageScreen uses structured interviews anchored to defined rubrics. While conversations adapt naturally in response to candidates, every evaluation is grounded in the same criteria for the role. Candidates are assessed against the rubric, not against each other.

Interview questions are generated dynamically based on role-specific guides, categories, and guardrails configured per Sage. Customers influence what is evaluated, but they do not handcraft question lists or adjust scoring on the fly. This prevents drift, coaching artifacts, and uneven application of criteria.

Once an evaluation is produced, it is not normalized, adjusted, or rescored. There is no curve, no relative ranking, and no post-processing designed to shape outcomes. What the rubric yields is what the system reports.

The result reflects whether a candidate Meets Expectations or Does Not Meet Expectations for that role, based on that interview alone. This framing is deliberate. It avoids comparative judgments and keeps focus on role fit rather than competition.

Fairness improves when the same standards are applied consistently and explained clearly. That is the discipline we enforce.

Security Fundamentals

Security is not a feature set. It is baseline hygiene.

Access to SageScreen is governed by role-based access controls and multi-factor authentication. Users only see what they are permitted to see, and nothing more.

Data is encrypted in transit and at rest using provider-managed encryption. Sensitive access paths are gated through hashed controls to prevent unauthorized traversal or exposure.

Customer data is logically isolated. One customer’s data is not visible to another, and systems are designed to prevent accidental cross-access by default.

Operational monitoring is continuous. We track files, network activity, ports, databases, sessions, transactions, events, and credit and ledger activity. These signals exist to detect misuse, failure, or unexpected behavior, not to observe candidates.

There are no buzzwords here because none are needed. These are fundamentals. If a hiring system cannot get these right quietly and consistently, it should not be trusted with real people’s careers.

Data Retention and Deletion

We keep data only as long as it serves a clear purpose.

Personal candidate data, including interview transcripts and evaluation artifacts, is retained for up to four years by default. This window exists to support hiring audits, dispute resolution, and longitudinal review by customers who need defensibility over time.

Aggregated and sanitized data, stripped of personal identifiers, may be retained indefinitely. This data is used to understand system performance and improve reliability without exposing individuals.

Customers can initiate deletion by cancelling their account and explicitly requesting data removal. Candidates may also request deletion through our contact process. These requests are honored within 30 days.

Sensitive artifacts, such as images used for integrity checks, follow shorter retention windows and are automatically deleted in batch processes.

Retention is not about convenience. It is about balancing accountability with restraint.

Compliance Posture

We do not treat compliance as a badge to display. We treat it as an ongoing obligation.

Today, SageScreen is not formally certified under specific regulatory frameworks. We are actively aligning our practices with applicable privacy and data protection laws, including GDPR principles and relevant US regulations, but we do not claim compliance we have not earned.

We intentionally avoid collecting or processing data that would require heightened regulatory exposure when it is not essential to hiring evaluation. This includes biometric analysis, psychological inference, and sensitive personal profiling.

We do not rely on subprocessors that materially change our compliance posture, and we are cautious about introducing dependencies that would.

Compliance is not a finish line. It is a process that evolves as the product, the law, and expectations change. We choose to move deliberately rather than prematurely declaring victory.

Transparency Over Theater

Trust erodes faster through opacity than imperfection.

Customers can see high-level evaluation results and control key aspects of how Sages are configured during creation. The intent is not to expose internal mechanics, but to make outcomes understandable and reviewable by humans.

Candidates are informed up front about what to expect. They are told that AI is involved, that camera and audio checks are used for integrity, and how the interview process works. There are no hidden steps or undisclosed evaluations.

Explanations are delivered as structured summaries and plain-language context, not raw system internals. We aim for clarity without overwhelming users with technical detail.

Some aspects of the system remain intentionally opaque. This includes proprietary methods that, if disclosed fully, would make misuse easier or undermine fairness. Transparency does not mean handing out the keys to game the process.

We share what matters. We protect what must be protected. That balance is deliberate.

How We Evolve Responsibly

Change in hiring systems should never be silent.

SageScreen evolves through versioned releases, feature flags, and controlled rollouts. Changes to behavior are intentional, documented, and reversible.

Customers can access release notes through our public changelog, and they can choose whether to follow updates closely. We do not silently alter system behavior in ways that would affect outcomes without notice.

Rollback is supported. If a change introduces unexpected behavior, it can be reversed.

Responsible evolution means moving forward without breaking trust. Stability, clarity, and communication matter more than novelty.

A Final Word

If this all feels a little slower, a little more deliberate, that’s on purpose.

Hiring doesn’t need more spectacle. It doesn’t need mystery, hype, or systems that move so fast no one can explain what just happened. It needs calm. It needs clarity. It needs tools that help people think better, not decide for them.

SageScreen was built with the assumption that every candidate is a real person, and every hiring decision carries weight. Careers bend around these moments. So do teams, cultures, and companies. That’s not a place for secrecy or shortcuts.

Around a fire, with a glass in hand, this is the simple version:
If a system needs to hide how it works to be effective, it doesn’t belong in hiring.

We believe trust is earned by showing your work, admitting your limits, and leaving judgment where it belongs: with people. AI should make the process calmer, clearer, and more humane. When it does anything else, it’s doing too much.

That’s the line we hold. And we intend to keep holding it.