Stories & results

Different people.
Meaningful progress.

These are experiences from earlier Bolt Practice learning and Think With Me. They show the variety of work people bring, the part AI can play and why human judgement still matters.

These are individual accounts from earlier Bolt Practice programmes. They illustrate different starting points, applications and limits. Results depend on circumstances, effort and judgement; they are not guaranteed.

Family and everyday life

From lookup questions to a calmer home.

Hannah Evans

Hannah began with occasional ChatGPT questions. During the Accelerator she built practical systems around the household routines that caused the most friction.

Meal planning started with the food her family actually eats, their weekly activities and the contents of the cupboards. A pantry record helped her see what was in stock, low or missing. She then used AI to help fill an online shopping basket, checking brands and purchases herself.

She also developed a home ledger for incoming emails, a school hub and a Christmas gift tracker. The work was grounded in the demands of three children at three schools, rather than a generic productivity exercise.

Hannah reported that the weekly shop took around five minutes to review, compared with about 45 minutes before. This was her own estimate. She had not established an overall financial saving: the main shop cost more, while top-up visits had reduced.

Not everything worked. A nutrition and fitness tracker still needed improvement. She described herself as a calmer parent while continuing to fact-check and make her own decisions.

“I do trust it more now, but I do fact-check and make sure that the decisions are my own.”

Hannah Evans, Accelerator participant account

Useful household systems, with time and money claims kept distinct.

About the evidence +

Participant account from the September 2026 Accelerator, documented 2 November 2026. Source: Accelerator participant account, November 2026. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Business and connected work

From looking things up to developing ideas.

Nicky Badenoch

Nicky was building Our Creative Edge with her business partner Tanya, alongside her own consultancy. She wanted AI to help develop her thinking and connect practical work.

She used AI to develop a complex business model, prepare a brand strategy and build a website. Context about the business and its values helped the different pieces fit together.

A reusable research skill produced brand briefings for outreach. A connected personal assistant helped organise diary entries, summarise dates and identify unanswered emails.

The systems still needed oversight. Calendar entries sometimes went into the wrong calendar; the website back end was still to be built. She identified connected working as an area to develop further.

The change was broader than a single output: she began asking whether repeated work should become a reusable skill. Her account describes development and unfinished work, rather than established business revenue.

“I feel like I've got my own ideas, but it helps me develop my own ideas.”

Nicky Badenoch, Accelerator participant account

A more connected way to develop and organise work, with her own ideas still central.

About the evidence +

Participant account from the September 2026 Accelerator, documented 2 November 2026. Source: Accelerator participant account, November 2026. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Professional practice and judgement

Choose the tool. Keep the final say.

Alli Gibbons

Alli runs a two-person professional practice. Her Accelerator work increasingly became a question of architecture: which tool to use, where information belongs and who decides.

She separated client-specific context from reusable methods, developed workshop materials and rebuilt a quiz used as a lead magnet. The remaining quizzes still needed work before an existing subscription could be cancelled.

For assessment marking, she marked the work herself first and then compared an independent AI assessment with her own. She said the process did not save time; its value was stronger reasoning and a clearer audit trail.

When her IT administrator objected to connecting another AI system to Microsoft 365, she took the concern seriously and explored an alternative route. Tool choice followed the task and the information boundary.

Her account includes continuing errors and limits in sharing. She retained responsibility for the final work and described checking it closely.

“It's my name on the work, so it has to be my work.”

Alli Gibbons, Accelerator participant account

More deliberate tool choices and more defensible professional judgement.

About the evidence +

Participant account from the September 2026 Accelerator, documented 2 November 2026. Source: Accelerator participant account, November 2026. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Family, work and career

One family. Three different starting points.

Caroline, Mike and Ashley Marsh

A care-home business, a bookkeeping process and a career in football. Caroline, Mike and Ashley brought different work to the same practice.

Caroline runs children’s care homes. Before asking AI to help, she decided what it must never see: the sensitive records of the children in her care. With that boundary clear, she explored meeting preparation, recurring communications and reusable templates. The starting point was her responsibility to people.

Mike turned to the bookkeeping: could imports and transaction classification become easier to manage? That was work under review, with a bookkeeper involved. The evidence does not establish a completed saving.

Ashley was pursuing football player care. He used AI to support dissertation research and targeted outreach. Earlier programme accounts report a first-class dissertation and roles with Swindon Town and West Ham. Steve’s update on 13 November 2026 reports paid full-time player-care employment at Swindon Town, alongside his West Ham work.

Their goals were different. What they shared was a way to work through them, with their own experience, effort and judgement still central. The accounts do not establish AI as the sole cause of the results.

“The impact it’s had on my life, my family and the business has been incredible.”

Caroline Marsh, quoted in Think With Me

Ashley’s progress into paid player care sits alongside practical changes in how the family approached work.

About the evidence +

Published account and participant reports, with Steve's November 2026 update. Source: Draw the Line, Chapter 11. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Career and learning

From a temporary contract to a permanent role.

Sam Regan

At nineteen, Sam wanted to make a six-month opportunity last. He began with the part he could control: becoming more useful, proactive and reliable.

Sam had started a temporary contract-support role with almost no Excel experience. There were stretches when he had not been given enough to do. He began using that time to learn a skill the job actually needed.

With his coach and AI, “get made permanent” became a question he could act on each day: what could he do now to be more proactive, more reliable and more useful? The change was in what he did with the time in front of him.

Two weeks before programme graduation, his employer made the role permanent. Sam credited AI with being a major part of his progress. His own work, coaching, the opportunity and the employer’s judgement mattered too.

He also built a “good enough rule” for difficult days: a realistic minimum that stopped one lapse turning into a whole day lost. It gave him a way back into practice when motivation faded.

His temporary role was made permanent two weeks before graduation, according to his published account.

About the evidence +

Participant report in the current book. Source: Good Enough, Chapter 8. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Business decisions and practical work

A year’s bank transactions. A cashbook in about two hours.

Ash Pattani

Ash said this annual job usually took him five days. This time, he worked through the bank statement with AI and had a cashbook ready for his accountants in about two hours.

Ash is a chartered and certified accountant with established gym and property businesses. At year end, his rental-property transactions had accumulated. He downloaded a full year’s bank statement and worked through it with AI.

“We had to dig around a bit,” he said. About two hours later, he had a cashbook ready to send to his accountants. The work still needed his knowledge and attention. The output was a cashbook, rather than finished accounts.

The five-day and two-hour figures are Ash’s own comparison. Neither attempt was independently timed or established as exactly like for like. It is a specific account of useful progress, rather than a general saving we can promise.

For two possible gym sites, he used AI differently: voice conversations gave him somewhere to talk through the financials and unfinished questions. He returned to the agents with clearer thinking. Both sites remained under consideration; the decision and responsibility stayed with him.

A cashbook ready for professional review, plus clearer thinking about two potential business commitments.

Ash is also an investor in the company behind Bolt Practice.

About the evidence +

Participant's own comparison; not independently timed. Source: Empty Chair, Chapter 4. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Judgement and business decisions

Eight quotations checked. Two remembered wrongly.

Tim Rees-Eggert

Tim left a difficult commercial meeting believing he knew what had been agreed. Checking his notes against the transcript changed his position.

Before the meeting, Tim used two role-defined AI agents to challenge his thinking from commercial and legal perspectives. He prepared, took notes and recorded the discussion.

Weeks later, he asked AI to compare his notes with the transcript, quotation by quotation: “Tell me where I was wrong.” Six important quotations held up. Two did not.

A longer working day had been discussed, rather than agreed. A contribution to protective-equipment costs had been agreed by his side alone. Tim corrected the basis for his next conversation instead of relying on agreements that had never been made.

The checking went both ways. When AI later produced a calculation from an earlier invoice draft, Tim caught the error by supplying the actual invoice. The workings were corrected before they went to the customer. The commercial claim had been accepted in principle but remained unpaid.

“My memory is not evidence.”

Tim, quoted in Think With Me

He corrected two important recollections and an AI calculation before relying on them in the commercial claim.

About the evidence +

Attributed account in the current book. Source: The Other Version, Chapter 6. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Professional judgement and teamwork

Eight roles made coherent. Workshop ideas turned into a working draft.

David Klaasen and Alli Gibbons

David and Alli brought years of leadership experience. AI helped them organise the detail into work a client could examine, challenge and use.

David had eight job descriptions developed over about eight years. Individually they worked. Together, the progression had gaps and repeated responsibilities. He used AI to build a grid across the roles so he could see what changed, what repeated and what was missing.

He took the structure back to the client for approval, then developed the detail one layer at a time. AI found things David had missed. David found suggestions that sounded plausible but made no sense. He checked every line and used a fresh conversation to review the result against the project criteria.

For a senior-team workshop, Alli had to redesign the session when one participant could only join remotely. The team contributed to a digital whiteboard. The next morning, she used AI to organise those contributions into a six-page draft leadership framework.

Alli estimated about ten minutes for that first draft. The preparation, workshop and later review sit outside the figure. The ideas belonged to the people in the room; the draft gave them something concrete to examine and improve.

“I had to read every single line”

David Klaasen, quoted in Think With Me

A client-approved role progression and a six-page first draft built from a team’s own contributions.

About the evidence +

Attributed professional accounts in the current book. Source: Every Single Line, Chapter 9. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

Business systems and practical creation

From an operations manual to a working portal prototype.

Anna Rees-Eggert and Bethany Witherington

Bethany demonstrated an onboarding hub and operations manual for Skin Within. Her knowledge of the day-to-day work shaped what the prototype needed to do.

A new franchisee would see two choices: the onboarding hub and the operations manual. Behind them were tasks, forms, kit and standards. Staff could see progress; established clinics could bypass onboarding they no longer needed.

Bethany worked with AI to turn knowledge already held in the business into an interactive prototype. She estimated no more than two hours of active work so far, spread across about a week. That was her estimate at the demonstration.

The portal was still a prototype. Review and finishing remained, the existing paid service was still in place, and no saving had been realised. What they could assess was more concrete than an idea: a working demonstration of how the process might become easier to use.

A demonstrated portal prototype, with review, finishing and adoption still ahead.

Anna is also an investor in Athena Veritas, the company behind Bolt Practice.

About the evidence +

Demonstrated prototype and participant time estimate. Source: Behind the Screen, Chapter 10. These are individual accounts; results are not guaranteed.

Explore the complimentary book ↓

From the foreword to Think With Me

“Not to hand over my thinking
but to improve it.”

Kevin Cahill CBE

Reflecting on learning conversational AI with Jude Bolton.Get the free book ↓

Bolt Accelerator · The 12-week coached programme

Build a practice
you can use.

Bring a real goal and a willingness to practise. We will help you decide whether this is the right place to work on it.

Begins 29 January 2027 · Bournemouth · £3,500 + VAT per participant
Values
Judgement
Thinking
Habits