People First

Fill in the sheets, then save a PDF with your answers. Your answers stay in this tab until you save them.

Tomer Gal
TOMERWAVE / FIELDWORK
01 / The Friction Map

The Friction Map

Where does the work get stuck, and who feels it?

Choose one process and walk through it with leaders and the people doing the work. Record what actually happens, separate evidence from assumptions, and agree on where to begin.

Questions to work through

  1. What should this process make possible, and what matters to the people doing it?
  2. Where do people wait, enter information twice or send work back for corrections?
  3. How often does it happen, and who gets held up?
  4. What have we observed or measured, and what is still an assumption?
  5. Where should we start, considering impact, effort and the team's readiness?

Fill in together

Start and finish, steps, people and systems

Where work gets stuck; frequency, time, errors; sources and missing data

Verified causes, assumptions, concerns and what to preserve

First improvement, why it matters, owner and review date

Leave with a mapped process, evidence and a shared decision on where to start.

Tomer Gal
TOMERWAVE / FIELDWORK
02 / Before the Bot

Before the Bot

Do we need this task at all?

Start with the outcome and consider the options: remove, simplify, use an existing solution, automate or build. Choose based on the need and the ongoing cost. AI is an option, not the starting point.

Questions to work through

  1. Who uses the result, and what would happen if we stopped doing this task?
  2. Which steps could we remove or simplify without losing what matters?
  3. Could a tool or capability we already have be enough?
  4. What needs human judgment, and where could clear rules or AI help?
  5. What will it take to run and maintain, and what small test would help us choose?

Fill in together

Who benefits; the outcome and what must stay

Remove, simplify, use what exists, automate or build; benefits and limits

Benefit, total cost, privacy, risk, maintenance and owner

What to try, who will check, when, and what would change our decision

Leave with a reasoned choice and a small test before committing to a build.

Tomer Gal
TOMERWAVE / FIELDWORK
03 / The Human Line

The Human Line

Where does a person need to stay involved?

Work through each action. Decide what can run automatically, where the system assists and a person checks and decides, and what stays entirely human. Set the boundary around the cost of error, context and relationships worth protecting.

Questions to work through

  1. What harm could an error cause, and how can someone correct or challenge it?
  2. Where do empathy, context or human judgment matter?
  3. What may the system do, and what needs review and approval before action?
  4. When must it stop and hand over, and to whom?
  5. What information and training does the reviewer need, and how will we check that review actually happens?

Fill in together

Who is affected, cost of error, sensitive information and human contact

Automatic / assistance with human approval / human-only; allowed and prohibited actions

Clear triggers, named owner and backup, and a manual fallback

Review evidence, decision records, an exception scenario and a review date

Leave with decision boundaries, a named owner and a clear path for exceptions.

Tomer Gal
TOMERWAVE / FIELDWORK
04 / Time for What?

Time for What?

What will the time we free up make possible?

Before making a change, agree on what the recovered time should make possible. Afterward, check the actual saving, including review and maintenance, and what changed for people. A lighter workload is a worthwhile outcome too.

Questions to work through

  1. How much time and effort does this task take today, and how do we know?
  2. What would the team like to do with the time recovered?
  3. How much time will review, corrections and maintenance still take?
  4. Was time actually freed up, or did the work move to someone else?
  5. What improved in service or workload, and what do the people affected say?

Fill in together

Hours per period at comparable workload; data source; review, corrections and maintenance

The team's chosen purpose, owner and how time will actually be allocated

Net time, service quality, workload, usage and feedback

Date, participants, evidence and what to change if the benefit did not materialize

Leave with a purpose for recovered time and a way to check the real benefit.