Use cases Approach Services About Contact Book a call
Fractional AI advisory

Make good AI use common practice.

Common Practice helps lean executive teams turn recurring, high-value knowledge work into AI-assisted workflows their staff can actually run. We find where AI is worth applying, redesign the workflow around real materials, build the first usable version, add review gates, and train the team to own it.

10+
years turning recurring operational work into workflows teams own.
Built inside
Goldman Sachs · Uber · Cruise · Credal
Clients served
Federal agencies · Fortune 50 · Nonprofits · SaaS
Patterns

A few examples of what this looks like in practice

01Investment sourcing
Before

Opportunities are tracked across inboxes, referrals, websites, spreadsheets, and individual judgment.

After

A structured screening workflow that captures source materials, applies consistent criteria, flags diligence gaps, and prepares the next human review.

02Executive and stakeholder prep
Before

A senior person pulls together background, relationship history, notes, risks, and open questions before every important meeting.

After

A repeatable prep workflow that drafts concise briefs from approved sources and highlights what needs human judgment.

03Recurring reporting
Before

Teams rebuild updates from raw notes, metrics, documents, and prior decks each month or quarter.

After

A source-grounded reporting workflow with reusable structure, review checkpoints, and clear ownership.

04Market and policy monitoring
Before

Relevant updates are scattered across newsletters, websites, filings, emails, and "did anyone see this?" messages.

After

A standing monitoring workflow that summarizes relevant changes, explains why they matter, and routes them to the right owner.

05Internal knowledge and decision support
Before

Staff rely on memory, Slack searches, old docs, and repeated questions to find how work gets done.

After

A maintained AI-assisted knowledge workflow that helps the team retrieve context, draft next steps, and know when to escalate.

How we work

Most AI projects fail on adoption. We start where the work actually breaks.

01 · Workflows before tools

The problem first

The hard part is defining the work step by step: mapping the realistic, efficient version of the workflow before any tool enters the picture, then documenting it clearly enough that the whole team can follow it. Once the workflow is that clear, the right tool is obvious.

02 · Honest scope

We'll tell you what you don't need

We know what's overbuilt, what's maintainable, and what's actually worth doing. We flag what's overbuilt before you pay for it.

03 · Implementation

Built and handed off

We scope a discrete piece of work, build it alongside your team, and hand it off in a state you can own. When we leave, the workflow stays and keeps running.

The hard part is rarely the tool.

Choosing the right workflow, designing the review gates, training the team, and making the new way of working stick — that is the work. Most engagements spend 90% of the time on the tool and ten percent on when to use it. We reverse that.

Fit

Who we work with

The practice is kept small and the relationships close. That means it is a fit for some teams and honestly not for others.

A good fit

  • Lean executive teams with recurring knowledge work that eats senior time
  • Organizations where leadership will be in the room, not just sponsoring from a distance
  • Teams that need better screening, reporting, research, documentation, or internal knowledge workflows
  • Low to moderate AI maturity, with a real operational problem to solve

Probably not a fit

  • Teams looking to outsource a custom AI build with no internal change
  • Anyone wanting a strategy deck with no implementation behind it
  • Organizations without leadership buy-in for the work
  • Projects chasing a demo rather than durable adoption
Services

Every engagement is scoped, outcome-defined, and handed off cleanly.

Most clients start with a workshop or move straight into an adoption sprint once we know where the value is.

Flagship Primary engagement · fixed scope

Adoption sprint

A fixed-scope, four-week engagement that opens with a workflow audit, builds the highest-leverage workflow on your real cases, and ends with your team owning it.

Book a workflow discovery call
How the sprint works · four weeks, not a slide deck
01
Week 0 · Workflow auditWe interview the team, observe how the work actually moves, review the tools already in use, and identify the workflow worth building first.
02
Weeks 1 to 2 · Build with youWe build the first workflow on your real work, with the right context files, quality checks, human review points, and owner handoff built in from the start.
03
Weeks 3 to 4 · Transfer ownershipYour team adapts the pattern to additional workflows. We facilitate, debug, and document until the workflow can survive without us.
À la carte
Entry point · flat fee

Executive workshop

A half or full day with your leadership team. We map where AI will move the needle and where it won't, and leave with a prioritized, honest list you can act on.

Flexible · four to six weeks

Team training series

A structured series that turns a workflow into repeatable team skill: hands-on sessions on your real cases, a written playbook, and review gates your staff run themselves. Runs alongside a sprint or on its own, scoped to the depth your team needs.

Limited availability · from Sep 2026

Operating partner

After a sprint, we stay embedded on a limited basis: implementing the next workflows, deepening team training, running office hours, and keeping the backlog moving.

On pricing.

Engagements are priced against the value of the problem being solved, typically 10 to 30 percent of its first-year cost to you. If we can't estimate the return together, the scope isn't ready yet. We work that out on the call.

About

I'm Sharon Park. Common Practice is a fractional AI lead for lean teams, turning recurring knowledge work into workflows their staff can run.

I have spent more than a decade inside operations where the work was complex, cross-functional, and high-stakes — from Goldman Sachs to Uber to Cruise to Credal AI.

At Credal, I was the first business hire and advised enterprise, government, and tech customers on getting AI use to show up in weekly work, not just individual experiments: where it could help, what needed governance, and what had to change in the workflow for the tool to matter.

Common Practice is built around that lesson. The hard part is rarely access to an AI tool. It is choosing the right workflow, designing the review gates, training the team, and making the new way of working stick.

I run a selective client load by design: senior-led work, direct implementation support, no bench of junior staff between us.

Book a workflow discovery call
Contact

Let's find out if there's a workflow worth building.

The best place to start is a 30-minute discovery call. Tell us where the work feels stuck and we'll tell you, honestly, whether we can help and what we would build first.

Book a workflow discovery call

Prefer email? Reach us at sharon@commonpractice.ai