automation strategy

I developed a user-centered, data-driven, design-first, 5+ year product strategy for the end-to-end Hyperscience workflow focused on strategic use of machine learning to minimize manual work.

impact

As a result of the strategy that I developed for Hyperscience, use-case implementation and model training times were reduced, allowing customers to onboard new use-cases quickly, with minimized manual work, increased accuracy and optimized automation in the 99% range.

This strategy additionally increased the total addressable market of use cases by enabling faster integration of new models - such as Hyperscience's proprietary VLM model ORCA.

role

product vision, product strategy, product discovery, concept design, user research & testing

team

Design, Product, Engineering

the research

what’s working? what’s not working?

I started out scoped to do a product and user feedback audit for just one specific product surface, but the step-wise dependencies of a complicated enterprise product meant the identified problem themes had compounding effects across the product.

My research and discovery illuminated 3 areas of strengths and friction:

- flexible, but unintuitive

- powerful, but requires subject matter expertise

- many task flows, many dead ends

All of problem areas I discovered share a core theme — lack of a coherent and consistent product strategy tying the experience together.

the challenge

why doesn’t it just work? machine learning is ✨magic✨, right?

Customers were saying the same thing over and over

"The product claims to use machine learning for automating processes, so why am I doing so many things manually?"

The Hyperscience product was using machine learning for very specific document processing and text-extraction tasks, but not applying that same technology across the end-to-end business process automation workflow.

the proposal

automate everything

What if we could make the end-to-end Hyperscience platform deliver on the promise of machine learning? What if we could make the entire product really learn? A product that already knows what you want to do, and the best way to do it?

I focused on injecting automation into every step of the Hyperscience workflow, and applying a model training framework I developed.

The proposed strategy excited Hyperscience's C-Suite, GTM teams, and customers alike with its user-centered approach that delivers what customers really want - maximized automation and reduced manual work.

This is like Steve Jobs with the iPad - it just works!

CEO

@ Hyperscience

I can sell this at the executive level tomorrow!

CTO

@ Hyperscience

This makes me want to break dance.

evp, global revenue

@ Hyperscience

“[Meg] spent months shaping our end-to-end automation strategy, grounding her ideas in insights from go-to-market teams and a deep analysis of customer tickets. The outcome was a systemic approach to document automation—an ambitious vision translated into pragmatic milestones that consistently delivered customer value.”

director of product

@ Hyperscience

The concepts and point of view demonstrated by [Meg] address fundamental gaps in both our competitiveness, as well as strategic issues preventing further expansion with customers.

The ability to automate all phases of the Hyperscience workflow will allow us to address more use-cases faster, increase consumption, and get to full utilization and upsells more successfully.

enterprise account executive

@ Hyperscience

WOW! Amazing. Simply awesome.

sr. director of customer experience

@ Hyperscience

Today feels like christmas

sr. manager, solutions architecture

@ Hyperscience

THIS IS INSANE

manager, solution architecture

@ Hyperscience

the principles