SW6 AI Studio

What would you like AI to do in your business?

Explore what AI could do for your business. Turn your AI idea into a working application.

We design and build custom AI for small businesses in London.

You might already have something in mind. You might want help seeing the possibilities.

Let’s find the best place to start together!

Let’s talk

From idea to AI application

  1. Explore

    Bring an idea, a process or a question.

  2. Define

    Choose one idea. Agree scope, price and team input.

  3. Build

    We design and build your application.

  4. Try

    Test on real work. Decide what comes next.

What could AI take on for you?

Insurance broking · 01 / 18All AI Stories

AI insurance conditions coordinator

Helps commercial insurance brokers turn insurers' requirements into practical client actions, then assemble the evidence needed to show what has been completed.

At each renewal or policy change, the application connects the insurer's requirements to the client's actual premises, equipment and working arrangements. It prepares an action plan showing what needs doing, who must confirm it and what evidence the broker needs.

For example, a wholesaler's proposed policy includes different security requirements for its shop and overflow warehouse. The application prepares separate instructions for the two sites, requests the relevant installer records and identifies an old certificate that relates to equipment no longer in use.

The broker checks the interpretation with the insurer and agrees the actions with the client. The application then coordinates the approved requests, assembles the returned evidence against each requirement and prepares the submission for the broker. When the insurer responds, it records the broker-confirmed outcome and sets up any recurring checks.

Commercial property · 02 / 18All AI Stories

AI space planner

Helps small commercial-property owners turn upcoming vacancies into space offers that account for existing occupiers, physical constraints and the work needed before someone can move in.

Whenever a unit is due to become available, the application combines measured plans, lease dates, shared facilities and current enquiries. It prepares workable ways to offer the space, together with the access arrangements and alterations each option would require.

For example, an owner has two adjacent units becoming vacant at different times. One enquirer needs a workshop immediately and more storage later. The application prepares a phased offer using the first unit now and the second later, identifies the proposed connecting doorway and checks how the arrangement affects an existing occupier's access.

The owner chooses the commercial approach; the relevant building professional checks physical feasibility and permissions, and the legal adviser checks the proposed terms. Following those decisions, the application prepares the offer pack, landlord's works brief and staged availability records. Confirmed changes to a lease date update the affected offer before it is issued.

Construction and fit-out · 03 / 18All AI Stories

AI site readiness coordinator

Helps refurbishment and fit-out contractors turn the coming week's programme into work packages that are ready to start, with the necessary access, materials and preceding work checked.

Each afternoon, the application checks the next few days' work against current drawings, delivery confirmations, site reports and signed inspection records. It prepares location-specific work packs and the actions needed to clear anything holding them up.

For example, a ceiling is due to close on Thursday. The delivery is confirmed, but the required inspection of the services above it has not been signed off. The application keeps that area out of the ready-to-start list, prepares the inspection request and finds another area where the same crew's prerequisites have been met.

The site manager confirms the physical situation, approves the inspection arrangements and chooses the work sequence. The application issues the approved crew packs and delivery instructions. When the inspection is signed off, it prepares the original area for release.

Trades and installation · 04 / 18All AI Stories

AI commissioning partner

Helps installation and maintenance firms test how connected equipment works together, prepare adjustments and produce a checked record of the finished system.

For each installation or scheduled recommissioning visit, the application builds a test sequence from the equipment actually fitted, its configuration, the engineer's measurements and the customer's operating requirements. It follows the results through the whole system rather than treating each component separately.

For example, a new heating controller must work with existing ventilation and room controls. Individual units pass their checks, but the combined system starts before the customer's agreed opening hours. The application traces the conflicting schedules, prepares revised settings and creates tests for the normal start, manual override and shutdown sequences.

The engineer checks the proposed settings and test method. The application prepares an approved configuration for the supported controller, or instructions where settings must be entered manually. After the engineer carries out the work, it compares the recorded results with the agreed requirements and prepares the as-installed settings and customer guide for sign-off.

Manufacturing · 05 / 18All AI Stories

AI production recovery planner

Helps small manufacturers turn inspection failures into a workable plan for recovering usable parts and remaking only what is necessary.

At each batch inspection, the application combines recorded defects, measured dimensions, material records, production drawings and outstanding orders. It works out which pieces could be reworked, which could serve another approved order and which must be remade.

For example, a furniture workshop finds that several panels have been cut short. The application checks whether they can become smaller parts already needed elsewhere, prepares revised cutting layouts and calculates the replacement material required for the original job. It keeps the finish, grain direction and customer specification attached to every proposed reuse.

The production manager checks the measurements, acceptable rework and delivery priorities. After approval, the application prepares revised job cards, material reservations and cutting files for the workshop's usual machine checks. It then uses the next inspection results to account for the recovered pieces and remaining replacements.

Wholesale and distribution · 06 / 18All AI Stories

AI supply substitution planner

Helps wholesalers complete customer orders when the requested products are unavailable by preparing checked alternatives that still meet the customer's requirements.

During each day's order review, the application brings together customer specifications, supplier catalogues, confirmed stock, pack sizes and delivery dates. It prepares alternative ways to fulfil short lines, including the effects on quantities, price and other items in the order.

For example, a restaurant-supply wholesaler cannot obtain the takeaway containers a customer ordered. An alternative container is available, but it needs a different lid and arrives in a different pack size. The application checks the recorded dimensions and intended use, calculates the complete container-and-lid quantities and prepares a revised offer showing the differences.

The sales manager checks suitability and obtains the customer's agreement. The application then reserves the approved combination, prepares the supplier orders and updates the warehouse picking instructions.

Retail and product brands · 07 / 18All AI Stories

AI product improvement manager

Helps independent retailers and product brands turn repeated customer difficulties into tested changes to the products, packaging or instructions they sell.

Each week, the application connects returns, support conversations and inspection findings to the exact product version and supplier batch. It prepares an improvement brief and a practical way to test whether the proposed change solves the problem.

For example, customers repeatedly return a flat-pack shelving unit after fitting one panel backwards. The application links their photographs to the assembly step, prepares clearer instructions and proposes a small change to the panel labels. It also drafts a supplier request for a labelled sample and prepares a trial assembly checklist.

The owner checks the evidence and approves a trial. Staff test the sample; the application records where the instructions worked or failed and revises the pack for review. Once approved, it updates the purchasing specification and prepares the matching product-page and support materials. It tracks which stock carries the new version so later returns can be compared fairly.

Restaurants and hospitality · 08 / 18All AI Stories

AI kitchen trial coordinator

Helps restaurants test practical changes to dishes and preparation, then turn the results into approved recipes and service instructions.

For each weekly menu review, the application uses sales, preparation times, ingredient costs, waste records and staff observations to prepare a small trial. It specifies what will change, what should stay consistent and what the team needs to record.

For example, a popular lunch dish creates a queue at one cooking station. The application proposes testing a different preparation sequence on comparable lunch services. It prepares the trial recipe card, batch quantities, station instructions and a simple record of waiting time, waste and customer feedback.

The chef checks food safety, allergens and whether the method is practical, then approves the trial. The application places the approved trial instructions in the kitchen workflow and connects the resulting sales and preparation records to that version. It compares the results, including differences in demand or staffing that could explain them. The chef decides whether to adopt, revise or abandon the change; the application prepares the corresponding recipe, purchasing and training updates.

Care and wellbeing · 09 / 18All AI Stories

AI home adaptation planner

Helps occupational therapy practices turn a client's everyday difficulties into practical home changes, equipment trials and instructions for using the agreed setup.

For each home assessment and follow-up, the application brings together the therapist's findings, the client's goals, measured room layouts and approved equipment options. It prepares adaptations around the activities the person wants to carry out, with a plan for trying the proposed arrangement.

For example, a client wants to prepare breakfast more independently. Using the therapist's recorded reach and movement requirements, the application prepares an arrangement for storage and equipment, checks it against the measured kitchen and creates a trial checklist. It identifies where an appliance cable or an opening cupboard would interfere with the proposed setup.

The therapist checks suitability with the client and approves the trial. The application prepares the equipment-loan request, setup instructions and illustrated practice guide. After the therapist observes the trial, it uses the recorded findings to revise the arrangement and prepare any agreed installation brief. The therapist approves the final plan; the application updates the client's guide and follow-up checks around what was actually fitted and found useful.

Education and training · 10 / 18All AI Stories

AI practice workshop builder

Helps training providers create working practice environments where learners can rehearse a client's real processes and receive feedback on the actions they took.

For each training cohort, the application turns the employer's approved procedures and anonymised examples into practical exercises inside a separate training environment. It creates the starting records, the events that occur during the exercise and the checks needed to assess the result.

For example, staff learning order administration receive a practice order followed by a customer's change of delivery address and a partial cancellation. The application creates the linked customer, stock and order records. It then checks whether the learner changed the right records, released the correct stock and left a consistent order behind.

The trainer reviews and tests the exercise before assigning it. After learners complete it, the application prepares evidence-based feedback and a new practice case for the steps they found difficult. The trainer checks disputed results and decides readiness to progress. Each exercise can be reset and replayed without affecting live customers or stock.

Publishing and media · 11 / 18All AI Stories

AI archive licensing producer

Helps publishers turn their existing articles, photographs and other material into complete packages that can be licensed to another business.

For each licensing enquiry, the application searches the publisher's archive, assembles a proposed package and connects every component to its recorded permissions, corrections and contributor terms. It prepares the work needed to make the whole package deliverable.

For example, a food publisher receives a request for a collection of recipes for a hotel group's digital guest guide. The text is available for reuse, but some photographs were licensed only for the original magazine. The application assembles a draft collection, proposes suitable cleared replacements and prepares permission requests for any essential material still missing.

The editor checks accuracy and suitability; the rights manager confirms permissions and commercial terms. After approval, the application prepares the licensed files in the customer's required format, with the correct credits, permitted uses and expiry records. It checks the delivered package against the approved component list and sets up the usage reporting required by the agreed deal.

Removals and logistics · 12 / 18All AI Stories

AI move planner

Helps removals businesses turn a checked survey into an item-by-item plan for packing, loading and delivering a move within the real access constraints.

For every booked move, the application combines the surveyed inventory with confirmed dimensions, vehicle capacities, collection access and the destination layout. It prepares how items should be protected, grouped, loaded and placed at the other end.

For example, an office is moving across London. Some furniture must go into storage, while the reception furniture must be unloaded first at a building with a short loading-bay slot. The application works out a proposed load arrangement, separates storage items and prepares labels linking each piece to its vehicle position and destination room.

The move supervisor checks handling methods, weight limits and access measurements, and confirms uncertain items onsite. Following approval, the application issues the packing and loading sheets. Crew scans record actual loading and delivery; discrepancies produce a revised placement or collection task for the supervisor to approve.

IT services · 13 / 18All AI Stories

AI system migration engineer

Helps IT service firms rehearse a client's move between business systems, repair migration problems and prepare a verified transfer for approval.

For each client migration, the application maps records, relationships, permissions and connected processes into a test copy of the destination system. It prepares conversion routines and checks whether the client's real work would still function after the move.

For example, a distributor is changing its order system. The application moves sample accounts, products and open orders, then rehearses an amendment and a return through the connected warehouse and accounts systems. The records balance, but the warehouse still expects an old product code. It prepares a corrected interface mapping and repeats the transaction to check the stock movement and credit as well as the order itself.

The engineer checks the mappings, test results and recovery plan with the client. The application repeats the rehearsal from a clean starting point and prepares the approved migration steps. At the agreed changeover, it runs only the authorised steps, reconciles the transferred records and stops for review if a required check fails.

Visitor attractions · 14 / 18All AI Stories

AI accessible visit coordinator

Helps visitor attractions prepare a visit around a group's stated access needs, with matching routes, visitor materials and arrangements for the staff who will host them.

For each group booking, the application combines the visitors' chosen preferences with the venue's verified access information, available sessions and current closures. It prepares a visit plan and adapts approved interpretation material for the group.

For example, a group asks for a step-free route, regular seated breaks and an alternative to a loud audiovisual room. The application prepares a route through suitable spaces, creates a quieter version of the room's approved story and proposes the necessary session reservations and staff meeting points.

The visitor-services manager checks that the venue can deliver the arrangements, and the organiser confirms the plan. The application then creates the approved visitor guide, booking updates and staff instructions. If a space closes before arrival, it prepares a replacement route and matching materials for approval.

Recycling and recovery · 15 / 18All AI Stories

AI material recovery planner

Helps recycling businesses turn inspected incoming material into processing instructions and batches that meet confirmed buyers' requirements.

Each day, the application uses intake weights, staff inspections, test results, available processing equipment and buyer specifications to prepare a recovery plan. It connects the proposed output of each processing step to a potential saleable batch.

For example, a plastics recycler receives mixed grades from several commercial customers. The application prepares a sorting and processing sequence, estimates the usable yield from the site's recorded results and identifies which inspected fractions could meet a buyer's specification. It keeps uncertain or contaminated material outside the proposed batch pending review.

The site manager checks the classification, permitted processing route and buyer requirements. Following approval, the application creates the processing job sheets and provisional batch records. Actual weights and quality tests then replace the estimates. It prepares the verified product specification and dispatch documents for approval, keeping each output traceable to the material that went into it.

Equipment hire · 16 / 18All AI Stories

AI equipment configuration engineer

Helps equipment-hire businesses turn a customer's requirements into a compatible working system, with the right components, setup instructions and checks before dispatch.

For each equipment enquiry, the application combines the intended use and confirmed venue details with the hire company's inventory, manufacturer specifications and tested equipment combinations. It designs how the items will connect and work together.

For example, a customer needs a camera, sound and display package for a live presentation. The application prepares the signal path, checks connection formats and cable runs, and identifies the adapters and backup items needed. If the preferred camera is unavailable, it recalculates the configuration around an alternative rather than swapping the camera alone.

The technical manager checks the design and approves the proposed kit. The application then reserves the items, produces labelled connection diagrams and generates a bench-test checklist. Confirmed test results are attached to the dispatch pack. If an item fails, it prepares a compatible replacement configuration for review before the equipment leaves.

Business marketplaces · 17 / 18All AI Stories

AI group order organiser

Helps business-to-business marketplaces combine smaller buyers' requirements into supplier orders large enough to be worthwhile, while keeping each buyer's quantities, deadlines and conditions clear.

During each weekly buying round, the application compares open buyer requests against suppliers' minimum quantities, specifications, pack sizes and delivery terms. It prepares potential group orders and the individual offer each buyer would need to accept.

For example, several independent cafés need the same type of unbranded packaging, but none meets the supplier's minimum order alone. The application checks the requested sizes and performance requirements, prepares a shared order with separate quantities and delivery costs, and identifies which requests cannot use the same product.

The marketplace operator checks the proposal, the supplier confirms capacity, and each buyer approves their own terms. The application records those commitments and prepares the supplier order only when the agreed conditions are met. If a buyer withdraws before confirmation, it recalculates the remaining offer and seeks renewed agreement where terms change.

Service networks · 18 / 18All AI Stories

AI service recovery coordinator

Helps managed service networks turn incomplete or unsatisfactory jobs into agreed corrective work, then check that the promised remedy has actually been delivered.

After each completed booking, the application compares the promised service with the provider's evidence and customer feedback. Where something needs correction, it prepares a specific remedy using the network's service standards and the customer's circumstances.

For example, a managed cleaning network receives photographs showing that part of an agreed end-of-tenancy clean was missed. The application prepares a room-specific return-visit brief, checks the customer's access window and sets out the evidence needed to demonstrate completion. It also prepares any proposed customer adjustment under the network's policy.

The operations manager checks what happened, agrees the remedy with the customer and authorises the provider's instructions. The application creates the corrective booking and gathers the required completion evidence. The manager resolves conflicting accounts and approves closure or any financial adjustment. The application then updates the customer record and prepares a targeted provider briefing where the same missed step has appeared across several jobs.

A small investment to get started

See what AI could do for your business with a budget you’re comfortable with. Know the price upfront, try it on real work, and decide what comes next.

A 15-MINUTE INTRODUCTORY CALL
Free
BUILD PROJECTS FROM
£500
INDICATIVE FIRST-PROJECT BUDGET
£2,500–£3,000

Business purchases only. Project prices are indicative; your written quote confirms the scope and total.

All prices include VAT where applicable.

Pricing information
Explore your first AI project

Working together

The person you meet is the person who builds your application. You work directly with me to explore the idea, agree what it should do and review it as it develops.

I built video games and AI products. Had been management consultant and business operator and always have this desire to make things work better. Started this project to hear people's ideas on making their businesses better with AI and hopefully help them succeed.

We are building custom AI-enabled software for businesses across industries. We bring that together with your team’s professional knowledge to shape an application around your firm.

Let’s talk

Tell me what you’d like AI to do—or bring a question. You don’t need a finished brief.

Choose how you’d like to start.

About on-site meetings. Every project starts with a visit. The £150 fee includes VAT where applicable and is paid upfront. The fee for a completed visit counts towards your first development project.