AI is changing the way we work.

Sign up for our newsletter and be the first to know about exclusive events, expert insight, and breakthrough research—delivered straight to your inbox.

Submit

Please share a few additional details to begin receiving the Valence newsletter

By clicking submit below, you consent to allow Valence to store and process the personal information submitted above to provide you the content requested.

Thank you! Your submission has been received!
Please close this window by clicking on it.
Oops! Something went wrong while submitting the form.

valence RESEARCH

The hard problems in AI coaching aren’t about the coaching.

They’re about the AI.

Any AI system can be trained on coaching frameworks.

The hard thing is everything underneath: the memory, continuity, and contextual understanding to navigate three dimensions simultaneously.

The three dimensions of AI coaching

Alignment across these layers is what makes it work

When the capabilities of large language models emerged, most companies asked a simple question:How can we add an AI interface on top of our existing expertise?Trust is the condition for coaching to work at all.

Valence focused on a deeper problem: building the memory, systems, continuity, and trust required to navigate all three dimensions at once.

Get the complexity of all three right, and you have something that can genuinely know a person, earn their trust, guide their journey, and help them be better at their job. That is the problem Valence was built to solve.

01

The individual

Their goals, history, working style, growth patterns, and what's happening in their life at work right now.

02

The work they’re in

Their team's dynamics, their relationships, the moments and pressures shaping their day.

03

The enterprise

Its culture, values, leadership frameworks, strategic priorities, and the context that makes coaching relevant rather than generic.

Getting the complexity of these three dimensions right is a fundamental research problem, and the one our team has been working on since 2022.

The Valence Broader Research Team

PhDs

10+

Pioneers across AI, engineering, and mathematics.

Papers

100+

Co-authored across the team

Key Topics

Conversational AI

Information Retrieval

Organizational Psychology

Applied Machine Learning

Learning

Nadia supports your skill-building and learning programs and plays a powerful role in creating an AI-enabled workforce.

Strategy

Give every employee a partner for goal-setting, working through companywide transformations, and instilling company frameworks and values.

Coaching

Scale coaching programs across your organization, free up capacity for your HRBPs, and and offer daily support for frontline employees.

Performance

Make  reviews more efficient, detailed, and productive, with support for giving and receiving feedback. Drive effectiveness improvements you can see and measure.

AI Coaching Perspectives

Perspectives

Whether it’s data from millions of conversations, rigorous studies, perspectives on applied AI, or conversations with the world’s leading thinkers, Valence is sharing our experience and views on the most important topics in AI coaching.

Valence Advisors

Trust is the condition for coaching to work at all.

From day one, we made ethical AI the most important investment we could make — and we were rewarded with the very first ISO 42001 certification in our category. ISO 42001 is the international standard for AI management systems — the highest bar set for responsible AI in the enterprise.

We are the only AI coaching platform to hold it.

Geoffrey Hinton

Nobel Prize in Physics (2024) · Turing Award (2018)

Godfather of modern deep learning.

Ethan Mollick

Associate Professor, The Wharton School, Co-Director, Generative AI Labs at Wharton

Author, Co-Intelligence: Living and Working with AI (NYT bestseller). Named one of TIME’s Most Influential People in AI, 2024.

Prasad Setty

Lecturer, Stanford Graduate School of Business · Former VP People Operations, Google, Founder, Google’s People Analytics team

Led Projects Oxygen and Aristotle, the most influential research ever produced on what makes great managers and great teams.

Reid Hoffman

Co-founder, LinkedIn · Co-founder, Inflection AI and Manas AI

One of the foremost thinkers on the intersection of technology, networks, and human potential.

The Research Domains

The frontier problems in applied AI thatValence has had to solve

Memory & Continuity

How does an AI system remember meaningfully across sessions, months, and career arcs — without creating privacy risk or contextual confusion? Valence built a proprietary memory and context engine that makes Nadia's coaching cumulative rather than episodic.

Context at Scale

Nadia holds three layers of context simultaneously: the individual, the team, and the organization. Synthesizing these without losing coherence, accuracy, or privacy required original systems research — not prompt engineering.

Safe Personalization

Personalization in an enterprise context carries stakes that consumer applications don't: legal exposure, reputational risk, psychological impact at scale. Valence has developed frameworks and technical guardrails for how AI coaching can be deeply personal without being intrusive — and rigorously compliant without becoming generic.

Coaching Continuity Across the Talent Lifecycle

No prior AI system had been built to follow someone through goal setting, mid-year review, skill development, and year-end evaluation — carrying coaching context forward at each moment. Valence designed and validated this continuity across 100+ enterprise deployments.

Learning

Nadia supports your skill-building and learning programs and plays a powerful role in creating an AI-enabled workforce.

Strategy

Give every employee a partner for goal-setting, working through companywide transformations, and instilling company frameworks and values.

Coaching

Scale coaching programs across your organization, free up capacity for your HRBPs, and and offer daily support for frontline employees.

Performance

Make  reviews more efficient, detailed, and productive, with support for giving and receiving feedback. Drive effectiveness improvements you can see and measure.

Foundational Choices

Ethical AI isn’t a feature. It’s a foundation.

Trust is the condition for coaching to work at all.

From day one, we made ethical AI the most important investment we could make — and we were rewarded with the very first ISO 42001 certification in our category. ISO 42001 is the international standard for AI management systems — the highest bar set for responsible AI in the enterprise.

We are the only AI coaching platform to hold it.

SOC 2 Type 2 text on a light background.
Text displaying 'ISO 27001' in a simple, modern font on a light rectangular background.

Zero failed security reviews across 100+ enterprise deployments.

We don’t train on user data.

Valence chose not to train its own foundation models, and prohibits its AI partners from using user data for training. Every coaching conversation remains within the user’s account. This is not a legal hedge. It is a research and ethical position: the trust that makes coaching effective depends on privacy being unconditional, not contingent.

We build on the world’s best AI.

By partnering with leading foundation model providers rather than training our own, Nadia benefits from the most capable AI available, while Valence’s research is focused entirely on the layer that makes coaching work: memory, context, continuity, and personalization.