Manas Meena
Manas Meena
Manas Meena

Case Study 02 | AI-Native Vertical & Agentic Patterns

Building the first AI-native vertical on IES and the agentic patterns the platform now composes from.

Leading design for
project-based businesses
on IES — from a portfolio of features to the architectural
foundation of unified job costing.

Role

Design Lead

Year

August 2024 - October 2025

For 15 months I led design across the portfolio that made IES viable for project-based businesses. AI-powered Project Budgets, Profitability, Milestones, 20+ reports, and Tasks as a shared pattern across six product surfaces.


The work landed at the AI inflection point. Leadership had committed to AI-first features across the platform but the design system had no AI patterns to compose against. I established the agentic interaction patterns through shipped product work. From inline AI assistance to fully done-for-you workflows with human sign-off, working directly with dev teams to build them as a self-serve system that developers on other verticals could inherit without us in the room.

The portfolio's success earned the trust to drive a structural change and set the foundation for every business vertical IES serves. And it's why I was brought to the manufacturing team next, to deliver the same class of business vertical in a fraction of the time.

Scope of Work

Strategy
Research
AI-first interaction patterns
Cross-team partnership
Ecosystem pattern design
Strategy
Research
AI-first interaction patterns
Cross-team partnership
Ecosystem pattern design

How i navigated this project

02

Patterns over stand-alone features

Every AI interaction I designed had to answer two questions, does it work here, and does it scale beyond here? Patterns that proved out were formalized and handed upstream, engineered with dev teams for self-serve inheritance.

04

Trust before structure

The core design problem of agentic software isn't capability it's trust. How much should the AI do, and where does the human stay in control? I designed the platform's answer as a spectrum, matching agent autonomy to the stakes of the action. One bar held across every level: AI is invisible work inside the customer's existing workflow.

01

Features as instruments

Each launch (Budgets, Profitability, reports) was both a shipping commitment and a probe. 15+ lighthouse customer sessions across the portfolio, findings feeding prioritization directly.

03

The long game on architecture

Three separate feature discoveries pointed at the same structural gap in job costing. Instead of shipping a fourth feature, I made the case for the platform change.

Project details

The Problems

An AI-first mandate with zero patterns. Every team was about to invent its own AI interaction model, separate AI panels, inconsistent trust mechanics, agents behaving differently on every surface. The platform was one planning cycle away from fragmenting its AI experience before it existed.

Problem 01

The 9-month, $60K setup tax

Mid-market construction customers spent 3–9 months and $60K–$300K configuring our horizontal platform to fit their workflows. A barrier to adoption that handed deals to competitors with native construction functionality.

Problem 02

Company-level lock on job costing
Logistics center top view

Project details

The Opportunity

Establish the AI interaction patterns through real shipped work, calibrated to a spectrum of agent autonomy and then build them as self-serve components any team's developers can inherit

Opportunity 01

Setup tax → Industry- tailored defaults

View all your projects, tasks, and deadlines in one simple dashboard. No more scattered to-dos.

Opportunity 02

Company-level lock → Project-level flexibility on a unified platform

Break down big goals into smaller steps and keep momentum with clear priorities.

Opportunity 01

Setup tax → Industry- tailored defaults

View all your projects, tasks, and deadlines in one simple dashboard. No more scattered to-dos.

Opportunity 02

Company-level lock → Project-level flexibility on a unified platform

Break down big goals into smaller steps and keep momentum with clear priorities.

Opportunity 01

Setup tax → Industry- tailored defaults

View all your projects, tasks, and deadlines in one simple dashboard. No more scattered to-dos.

Opportunity 02

Company-level lock → Project-level flexibility on a unified platform

Break down big goals into smaller steps and keep momentum with clear priorities.

Logistics center top view

Project details

The Goal

Move IES from "horizontal platform with construction features" to "platform built around how project-based businesses actually work" and set the design direction that the next generation of vertical-specific work would extend.

Expected Outcomes

Outcome 01

Increased construction segment engagement

Delivered

Outcome 2

2X more profitable projects (users with profitability)

Delivered

Outcome 03

Flexible Project Costing as foundation for unified costing

Delivered

Project Details

The Process & Hard Design Calls

Some of the major inflection points throughout the project

  • 01

    AI native vision through reserach and discovery

    The research that shaped this work spanned months and multiple feature releases. Across the FY25 Projects portfolio I led research studies with 15+ lighthouse customer FMH, with findings going directly into prioritization decisions. The insight that anchored the AI native industry vertical approach to IES came from this work.

  • 02

    The strategic call - Autonomy matched to stakes

  • 03

    Designing AI-first experiences as patterns

Project details

Business Impact

67.0K+

Project budgets in 6 weeks

10.0K+

Project reports viewed

57%

Construction segment engagement

8X

more profitable projects (users with profitability)

Partner Testimonials

"Manas does an excellent job of defining the long-term "North Star" for our products. He creates input goals that impact the wider function by visualizing a clear target state. He pushes me as a PM and our engineering partners to not just settle for the MVP, but to implement a path toward that ideal state."

Person profile picture.

Vinayak S.

Staff Product Manager, Intuit

"He has the drive to grow into a bold designer who not only builds what's possible — but redefines what's expected."

Person profile picture.

Shaily Shah

Design Manager, Intuit

Project details

Business Impact

67.0K+

Project budgets in 6 weeks

10.0K+

Project reports viewed

57%

Construction segment engagement

8X

more profitable projects (users with profitability)

Honest Reflection

Partial Wins

The costing convergence shipped, but the win is partial: no bidirectional switching, no multi-currency. The release is the foundation, not the destination  and the one-way constraint limits adoption for customers who want to try Detailed mode and back out. With another quarter, that's where I'd have pushed. The second reflection is about narrative. Across 15 months, work that individually looked like execution collectively amounted to direction-setting  and the partners closest to it read it that way, while my management chain didn't always. Some of that is on them; some is on me for not making the through-line legible at the moments that mattered. The lesson I've internalized: at senior levels, the work isn't the unit. The work plus the narrative is. I shipped the work, this page is me learning to ship the narrative.

More Projects

More Projects

AI tool : Prototyping Workbench

Ai infrastructure to bring Figma and production closer..

View Project

AI tool : Prototyping Workbench

Ai infrastructure to bring Figma and production closer..

View Project