Productiv

Designing an Enterprise AI Governance Tool

Turning an emerging AI risk into a new product for Productiv, from research to launch in four months.

Productiv

Designing an Enterprise AI Governance Tool

Turning an emerging AI risk into a new product for Productiv, from research to launch in four months.

Productiv

Designing an Enterprise AI Governance Tool

Turning an emerging AI risk into a new product for Productiv, from research to launch in four months.

The CHALLENGE

AI was spreading across SaaS faster than teams could track or govern it.

AI was rapidly embedding itself into the everyday SaaS stack, but most teams lacked a clear way to track where it lived or measure how it affected their operations. CIOs and IT leaders were caught in a balancing act - accelerating AI innovation while protecting the business from data exposure.

This presented a clear opportunity for Productiv. With an established foundation in SaaS management, extending our platform to help enterprises govern their AI portfolios was a natural next step.

77%

of employees say they are using AI tools without informing IT

68%

of IT leaders admin they don’t have visibility into AI usage in their org

RESULTS

Led the design of ProductivAI, helping Productiv move beyond SaaS management into AI governance.

0 → 1

New product launched in 6 months

$760K ARR

Generated within 6 months of launch

24 customers

Include UserTesting, JPI and more

Research

Understanding real world needs

To understand the current challenges, we spoke with CIOs and IT admins responsible for managing AI across their SaaS environments. These conversations surfaced insights that helped shape the product direction.

1

Unsanctioned AI use is the new shadow IT

Employees were adopting tools like Notion AI, Claude Code, and ChatGPT without any review. IT discovered most AI tools reactively - through audits, security incidents, or finance flagging an unknown expense.

2

No clarity on which vendors train on your data

Leaders couldn't easily determine whether vendor AI models trained on company inputs. ToS were long, buried, and changed without notice. This was the single highest-anxiety issue across every interview.

3

Policy exists on paper, not in practice

Some teams had AI usage policies in Confluence or a PDF somewhere. There was no systematic way to know whether apps met the stated policy.

4

Teams wanted to say yes to AI, safely

The goal wasn't to block AI adoption. Leaders wanted a way to evaluate and approve AI tools faster, with confidence.

5

IT admin as information bottleneck

Every question about app ownership, security review, or AI risk landed with Isaac. Without a system of record, answering took hours.

Research

Two distinct users, one shared problem

Research surfaced two core personas with meaningfully different maturity levels, goals, and frustrations - but the same underlying fear: that AI was moving faster than they could control.

Strategic Sarah

CIO / Director of IT, medium to large companies

Sarah is responsible for shaping AI governance while still enabling employees to adopt new tools. She needs to understand where AI exists across the portfolio and how it impacts the company.

Understand AI risk across the application portfolio

Create and enforce governance policies

Demonstrate ROI and accountability to leadership

IT Isaac

IT Admin / Manager, small to large companies

Isaac manages a growing software stack with limited time and incomplete visibility. He needs to stay ahead of AI risk and app sprawl without adding more manual work.

Know which apps and AI tools are being used

Identify unmanaged apps, ownership gaps, and AI-risk

Reduce manual work around access, reviews and policies

Illustration credits: Open Peeps by Pablo Stanley

process

Making ProductivAI a reality

How we went from an emerging customer problem to a shipped product.

Phase 1

Research

We spoke with CIOs and IT leaders to understand how AI risk was showing up across SaaS portfolios.

Interviews

Surveys

Phase 1

Research

We spoke with CIOs and IT leaders to understand how AI risk was showing up across SaaS portfolios.

Interviews

Surveys

Phase 1

Research

We spoke with CIOs and IT leaders to understand how AI risk was showing up across SaaS portfolios.

Interviews

Surveys

Phase 2

Synthesis

Partnered with the PM to identify which AI governance problems mattered most to IT teams.

Value Proposition Canvas

JTBD

Phase 2

Synthesis

Partnered with the PM to identify which AI governance problems mattered most to IT teams.

Value Proposition Canvas

JTBD

Phase 2

Synthesis

Partnered with the PM to identify which AI governance problems mattered most to IT teams.

Value Proposition Canvas

JTBD

Phase 3

Shaping

Simplified the legacy product and shaped the new SKU around visibility, risk and action.

Prioritization exercise

Legacy feature audit

Phase 3

Shaping

Simplified the legacy product and shaped the new SKU around visibility, risk and action.

Prioritization exercise

Legacy feature audit

Phase 3

Shaping

Simplified the legacy product and shaped the new SKU around visibility, risk and action.

Prioritization exercise

Legacy feature audit

Phase 4

Design

Designed the core workflows, refined the experience, and supported the launch narrative with product and marketing.

Iterative validation

GTM Launch

Phase 4

Design

Designed the core workflows, refined the experience, and supported the launch narrative with product and marketing.

Iterative validation

GTM Launch

Phase 4

Design

Designed the core workflows, refined the experience, and supported the launch narrative with product and marketing.

Iterative validation

GTM Launch

DESIGN

Designing a focused AI governance experience

We transformed the strongest parts of Productiv’s existing platform into a simpler, AI-focused experience for governing apps across the SaaS portfolio.

Dashboard surfaces immediate risks and provides full visibility into their portfolio

From portfolio overview to granular detail - users can move between both with ease

AI automatically surfaces what reviewers need to know before they even ask

Teams can define their own fields to surface the portfolio insights that matter to them

A simplified status model that was driven by actionability

Built a realistic ProductivAI prototype to support Sales demos ahead of launch

A simplified status model that was driven by actionability

Designed launch pages in collaboration with Marketing and Product

Example of a finalized contract

© 2026 Lakshman. Designed in Figma and built using Framer.

© 2024 Lakshman.

Designed in Figma and built using Framer.

© 2026 Lakshman. Designed in Figma and built using Framer.