top of page

Contact
Artemisia de Vine

For media interest,
academic collaborations,
licensing inquiries,
or high-level mentorship,
you’re welcome to reach out directly.

Thanks for submitting!

deVinery Parallax Paradox
Case Studies

deVinery Paradox Parallax

Telecommunications Company Case Study

​

 

Purpose of this document

 

This is a working test and case study of the prototype deVinery Paradox Parallax (DPP). The raw conversation comes first so the reader encounters the ordinary situation before being given an interpretation. The later sections preserve the conventional analyses already developed for this case, then apply DPP and compare what becomes visible.

 

The case is deliberately mundane: a self-employed customer asks a telecommunications company for short-term payment assistance while waiting for a client to pay an invoice. That ordinariness is part of its value. The interaction can be recognised from the customer, customer-service officer and company positions without requiring specialist knowledge.

 

The primary source is a sequence of 22 screenshots. Names, company identity and the customer's mobile number have been anonymised. Spelling, grammar, repetition and awkward phrasing in the messages are otherwise preserved rather than tidied. The screenshots remain the source of record.​​

​

​

 

What to look out for in the below transcript

​

​

Each corner is trying to solve a different version of the problem to achieve the same overall outcome.

​

One corner thinks this is the problem.
Another corner thinks that is the problem.
The third corner thinks something else is the problem.

​

The Parallax reveals what the actual problem is, and once the target changes, new options become visible. The possibility space increases. 

​

As you read the dialogue in the telecommunications company case-study, see if you can notice:

​

Customer: “The problem is that I can’t give you an exact payment date without lying.”

​

Staff/system: “The problem is that I need an exact date to process the assistance.”

​

Company: “The problem is that we need predictable payment and manageable risk.”

 

 

Then the Parallax reveals:

​

The actual problem is not the date. It is that the system has confused one administrative strategy with the outcome it is trying to achieve. It thought that it had weighted the importance of each corner equally but in fact had pre-selected the customer as the least important factor. This resulted in losing the customer. 

​

Once that becomes visible, the possibility-space opens.
 

​​

​

 

PART I — ANONYMISED TRANSCRIPT

 

TELECOMMUNICATIONS COMPANY ASSISTANT (BOT):

 

I'm [Telecommunications Company] Assistant – a bot here to answer your questions.

 

Some of my responses will be AI-generated. Generative AI is experimental and isn't perfect. Check our website for the most accurate and complete information.

 

If I can't answer your question, I'll transfer you to a human expert.

 

How can I help?

 

CUSTOMER:

 

I'm self employed and waiting for my client to pay their bill. Please give me a payment extension until the end of the month. I expect the money by then.

 

TELECOMMUNICATIONS COMPANY ASSISTANT (BOT):

 

I'll get one of our experts to help you with this shortly.

 

You don't have to wait around – you can close the app. They'll reply to you as soon as they can. Make sure you've got push notifications on, so you'll know when they reply.

 

To turn them on, select 'Profile' in the app, then 'App settings' and 'Push notifications'.

 

CUSTOMER SERVICE OFFICER:

 

Hello [Customer] (please let me know if you prefer to be called something else), you are now messaging the Payment Assistance Team. My name is [Customer Service Officer], and I’m here to help you. I understand you have a question about a payment extension, and I can assist you with that.

 

CUSTOMER:

 

Thank you

 

CUSTOMER:

 

I’m self employed and waiting for my client to pay their bill. Please give me a payment extension until the end of the month. I expect the money by then.

 

CUSTOMER SERVICE OFFICER:

 

I understand that managing finances can be challenging at times. To assist you better, could you please answer these questions for me?

 

1. Can you confirm your current source of income? (Full-time, Part-time, Casual, Centrelink)

2. Will there be any challenges in paying future bills?

3. Is this a one-off situation or a request for payment assistance?

 

CUSTOMER:

 

Did you bother to read what I said?

 

CUSTOMER SERVICE OFFICER:

 

Yes, I did what you said, and hence I have sent the above questions.

 

CUSTOMER SERVICE OFFICER:

 

To assist you better, could you please answer these questions for me?

 

1. Can you confirm your current source of income? (Full-time, Part-time, Casual, Centrelink)

2. Will there be any challenges in paying future bills?

3. Is this a one-off situation or a request for payment assistance?

 

CUSTOMER:

 

I’m self employed and waiting for my client to pay their bill. Please give me a payment extension until the end of the month. I expect the money by then.

 

CUSTOMER SERVICE OFFICER:

 

Could you please provide the number?

 

CUSTOMER:

 

What number?

 

CUSTOMER SERVICE OFFICER:

 

Could you please confirm your mobile number? and what number means your phone number

 

CUSTOMER:

 

You’ve asked questions s I have no way of answering without becoming g psychic. I have already told you my situation. [MOBILE NUMBER PROVIDED]

 

CUSTOMER SERVICE OFFICER:

 

Sure. I’ve checked your account, and I see that your bill is $74.00. Is that correct?

 

CUSTOMER:

 

No it’s only $44 this month because I used my points to pay some of it.

 

CUSTOMER SERVICE OFFICER:

 

Thank you for sharing this with me. You’re connected to the right person, so there’s no need to worry. I’ll make sure this issue is fully resolved.

 

Could you please let me know when you’ll be able to make the payment?

 

CUSTOMER:

 

I’m self employed and waiting for my client to pay their bill. Please give me a payment extension until the end of the month. I expect the money by then.

 

CUSTOMER:

 

Can you tell me why you are asking questions I’ve already answered repeatedly?

 

CUSTOMER SERVICE OFFICER:

 

I got you. But you don't have to send the same thing again. I was looking for the date you will be sending. Anyways, I can see that your anniversary billing date is the 8th of each month, if you’re experiencing any difficulties in making payments on time? If so, once you’ve cleared your current bill, you can look into moving you to invoice billing. This would allow you to make manual payments each month, giving you more flexibility to pay according to your situation.

 

CUSTOMER SERVICE OFFICER:

 

Please not nothing wrong in asking questions, as were helping you here.

 

Since you are currently on an Upfront plan, payments can only be deferred for a maximum of 10 days. Since the extension date has already been exceeded, there is no extension option available for this current bill. However, as a one-time exception, I can offer you a one-time payment relief by waiving this month's plan fee until 8 September 2026, so you can pay the upcoming bill.

 

Please note that during this relief period, your service will remain active without any restrictions. However, you will not be able to add new services or upgrade your existing ones. We want to ensure you can get your account back on track without added stress. If you have a device contract, it will be extended by one month.

 

With this information, would you like payment extension applied?

 

CUSTOMER:

 

Yes

 

CUSTOMER SERVICE OFFICER:

 

Now that we’ve resolved your enquiry, I’d like to do a quick account health check for you to ensure everything is updated and meets your expectations. If anything needs to be changed or added, just let me know. Would that be okay?

 

CUSTOMER SERVICE OFFICER:

 

Please let me know if you need any more help.

 

CUSTOMER:

 

Can you offer a cheaper plan? If not I will likely change companies

 

CUSTOMER:

 

My phone is not on a contract.

 

CUSTOMER SERVICE OFFICER:

 

Yes, that's correct. Upon checking your account, you're on the [Telecommunications Company] Upfront Mobile Plan - Basic. This is a basic plan, so there is no option to downgrade. If you're looking for alternative options, That's completely up to your choice.

 

CUSTOMER SERVICE OFFICER:

 

Is there anything else I can assist you with?

 

CUSTOMER:

 

No.

telecom case study table.png

PART II — CONVENTIONAL ANALYSIS BASELINE

 

This section preserves the baseline question: what would established customer-service, operations and organisational lenses see in this interaction before DPP is applied? The purpose is not to make conventional methods stupid. It is to show what they genuinely detect, what they would normally try to repair, and then compare the information available after a different measurement instrument is introduced.

​

Existing lenses already identified for this specimen include:

 

• CSAT, NPS and Customer Effort Score

• First Contact Resolution

• Average Handle Time

• Quality-assurance scorecards

• Root Cause Analysis / Five Whys

• Customer Journey Mapping / Voice of Customer

• Lean / Six Sigma

• Agent performance and active-listening assessment

• Payment-assistance / hardship-policy compliance

• Churn and retention analysis

• Staff burnout and engagement

• Chatbot, backend and workflow integration

 

A conventional analysis can plausibly identify high customer effort, repeated questioning, poor active listening, a rigid verification or hardship-assistance script, mismatches between customer information and available response categories, possible compliance or fraud-prevention requirements, legacy-system constraints, chatbot/backend integration problems, de-escalation failures and retention risk.

 

Likely conventional remedies include better training, better active listening, revised scripts, clearer questions, better routing, better chatbot or backend integration, improved exception handling, process redesign, quality assurance, staff support and retention intervention.

 

Blame or causal attention may move among the customer-service officer, the customer, the process or system, training, policy, technology and resourcing. These are real dimensions of the interaction. DPP is not being proposed because those dimensions are imaginary or useless.

 

The comparison question is narrower and more interesting: after those approaches have made the interaction legible in their own categories, is there structural information in the specimen that remains difficult to represent or measure?

 

 

PART III — PROTOTYPE deVINERY PARADOX PARALLAX TEST

 

Instrument under development

 

The deVinery Paradox Parallax is the instrument. Desire Signatures are what the instrument measures. This case uses a prototype DPP rather than claiming a finished or validated instrument.

 

Working test statement:

 

When we apply the prototype deVinery Paradox Parallax to this conversation, Desire Signatures become visible. The instrument is used to examine the residual between the triangulation actually occurring and the story or representation through which that triangulation is being processed.

 

The Three Paradoxes provide the relational coordinates:

 

1. Self vs Self

2. Self vs Other Self

3. Self vs Collective

 

A Desire Signature is the patterned way a differentiated self navigates those paradoxes. The same interaction can therefore be examined from the customer, customer-service officer and telecommunications-company positions rather than assuming that one position owns the whole description.

 

WORKING ANALYSIS SPACE — deliberately left open for collaborative revision

 

Customer

 

The customer states a concrete local reality: she is self-employed, is waiting for a client to pay an invoice, cannot know the exact date on which another person will make that payment, expects the money by the end of the month, and wants a payment extension until then.

 

A repeated behavioural move is visible in the transcript: rather than manufacture a more administratively convenient version of that reality, the customer repeatedly reintroduces the same information into the interaction.

 

Working formulation from the live analysis:

 

The customer is simultaneously defending her own corner and defending the isness of the local situation. She refuses to absorb the cost of the system's inability to represent her situation by inventing information reality does not contain.

 

A more accommodating customer could collapse the visible residual by translating herself into the available categories, supplying an approximate or invented date, or otherwise producing the administratively legible answer. In this specimen the customer does not do that. The refusal functions as a probe because the mismatch remains exposed rather than being repaired by the customer.

 

Customer-service officer

 

The officer is not simply the cause of the failure. She is another self inside the same triangulation. Her task is to help the customer while remaining inside the company's payment-assistance process. The transcript shows the available categories: Full-time, Part-time, Casual or Centrelink; whether future bills will be difficult; whether this is one-off assistance; and later, an exact payment date. The customer's actual answer does not fit those categories cleanly.

 

The officer therefore inherits the company's representation of the situation and has to enact it. She repeats questions whose answers are already present because the process requires information in a particular form. Her own move-space is constrained. If she departs from the process she may carry the employment or compliance risk; if she follows it, the customer carries the effort and frustration. This is why simply locating the problem in the officer's listening skills can misidentify the structural generator.

 

Telecommunications company / system

 

From the company's represented position, the arrangement appears reasonable. It is trying to keep the company viable, support staff to apply policy consistently, and assist customers experiencing payment difficulty. Existing processes also serve legitimate concerns such as payment risk, consistency, fraud prevention, compliance and operating at scale.

 

But under pressure, the enacted weighting visible in this interaction is different. Preserving the current process becomes structurally dominant. The customer's reality must be translated into the categories the process can recognise, while the officer has limited room to change those categories. The cost of maintaining that representation is displaced onto both customer and officer.

 

The company is not required to be 'lying' for this discrepancy to exist. Its stated map and its enacted weighting can simply be different. DPP is designed to make that difference measurable.

 

The key residual

 

The transcript contains a particularly useful moment:

 

“I got you. But you don't have to send the same thing again. I was looking for the date you will be sending.”

 

The customer has repeatedly supplied the reason that an exact payment date is unavailable. The process nevertheless continues to seek the missing date. The interaction therefore provides a candidate point at which represented reality and local isness can be compared directly.

 

The eventual outcome is also important: assistance is granted without the customer ever supplying the unknowable client-payment date. The practical outcome becomes available while the disputed piece of information remains unknowable.

 

The clearest specimen is the demand for a date. The customer has supplied the actual available information: payment depends on another person paying an invoice; she expects that to occur by the end of the month. An exact date is not information she possesses.

 

She could make the interaction easier by inventing greater certainty, choosing an answer shaped like the process expects, and becoming administratively legible. She refuses. Instead, she keeps reintroducing the excluded information.

 

That refusal is doing two things at once. It protects the customer's Self corner from carrying the entire cost of the mismatch, and it preserves representational fidelity to the actual situation. 'I am not going to carry this cost' and 'I am not going to invent a payment date that does not exist' are distinct moves occurring together.

 

This makes the customer's repetition analytically useful. A conventional lens can code it as frustration, escalation or difficult-customer behaviour. DPP asks what the repetition is doing structurally. Here it prevents the residual from disappearing. Reality keeps arriving in the interaction, while the system continues to request the version of reality its architecture knows how to process.

 

Then something revealing happens: the company grants assistance without the customer ever supplying the unknowable client-payment date. The demanded coordinate was not required for the eventual available outcome.

 

The residual is therefore not merely 'the customer was frustrated' or 'the officer failed to listen.' It is the measurable difference between the represented triangulation and the triangulation actually being enacted: which corners are weighted heavily, which are reduced, what information becomes inadmissible, what moves remain available, and who pays the cost of preserving the representation.

 

 

PART IV — INFORMATION DIFFERENCE

 

The difference can be understood through a three-legged stool.

 

Imagine a stool that keeps wobbling. Everyone examining the problem can see the same three legs, so naturally they assume one of the legs must be causing the instability. They trim one leg and test it again. It still wobbles. They wedge something under another. They reinforce the third. Each intervention is reasonable because the available model says that if a three-legged stool is unstable, something must be wrong with one or more of its legs.

 

Experts can become extraordinarily sophisticated at measuring those legs: their length, strength, material, angle and performance. They can develop better tools for trimming, reinforcing and compensating for them.

 

And the stool can still wobble.

 

Because nobody has measured how the stool is being weighted.

 

The three legs may be perfectly capable of supporting the stool. But if the person sitting on it places nearly all their weight over one corner, the forces running through the entire structure change. The resulting instability can look like a defect in the legs. Repeatedly modifying the legs will not reliably solve a problem generated by the distribution of weight across them.

 

That is the additional coordinate DPP introduces here. Customer, officer and company are not merely three separate factors to inspect and repair. They participate in one triangulated system, and each perceives that system from a situated position. Each has a representation of how Self, Other and Collective are weighted. The observable interaction lets us compare those representations with how the weight actually moves when the system is under pressure.

 

Existing instruments see many of the symptoms. DPP makes an additional structural relationship measurable between them.

 

Candidate distinction to test:

 

Conventional approaches can identify that the interaction is frustrating, repetitive, high-effort, poorly scripted or badly integrated and can recommend improvements to the process. DPP asks whether the process itself is enacting a patterned Desire Signature that systematically deforms the situation it is attempting to process, externalises the cost of maintaining that representation onto one or more corners, and constrains the move-space through which the problem can be solved.

​

​

Why locating the correct target matters: Reduced cost, time and harm at scale. 

​

The significance is not simply that this customer-service interaction could have gone more smoothly. It's that what everyone involved thought was happening, was not what was happening. You cannot fix what you cannot see exists. The problem gets mis-scaled, mis-categorised and mis-diagnosed. 

 

Once a system misidentifies the problem, everything it does next can be perfectly reasonable and still be aimed at the wrong target. That is where the cost compounds. Staff spend time repeating questions that cannot solve the problem.

 

Customers make repeat contacts, escalate complaints or leave. Managers respond with more training, better scripts, new procedures or new technology aimed at symptoms rather than the structural cause. Organisations spend money repairing things that were never the actual problem, while the real generator continues producing the same outcome.

​

At larger scales, the same distortion can mean misallocated investment, delayed projects, failed interventions, unnecessary restrictions, avoidable harm, displaced costs, policy fights, infrastructure mistakes and millions of dollars committed before anyone realises they were solving the wrong problem.

​

The Parallax creates a different category of decision:

​

Before deciding how to solve the problem, establish that you have correctly located the problem.

​

Once the target changes, decisions that were previously invisible can become available. The question is no longer limited to choosing between the existing options. We can ask what outcome those options were supposed to achieve, what information has been excluded, who or what has been made to carry the cost, and whether an entirely different route can achieve the desired outcome.

​

That is the economic jump.

 

PART V — DIRECT RELEVANCE TO THE AI AND DATA-CENTRE SUBMISSION

 

Same data. Additional coordinate. Different thing becomes visible.

 

Core bridge already identified:

 

“More data cannot compensate for a missing coordinate.”

 

This case offers a small, ordinary environment in which to test that proposition. The customer supplies the same relevant information repeatedly. The difficulty is therefore not simply absence of data. The research question is whether the representational system lacks a coordinate required to make the actual triangulation legible, and whether adding that coordinate changes what can be measured, predicted or redesigned.

 

The transcript has not changed. The customer has not supplied additional information. The officer has not supplied additional information. The company has not supplied additional information. The measurement apparatus changed.

 

This is the direct bridge to the AI and data-centre submission. Increasing data, compute or optimisation cannot reveal a relationship the measurement architecture has no coordinate for. This small case does not ask the reader to accept the larger deVinery claims in advance. It demonstrates a finite proposition on an ordinary dataset: existing approaches identify many genuine symptoms; introducing fixed relational coordinates and measuring their weighting makes an additional relationship between those symptoms visible.

 

That is the proposition the larger research programme proposes to test across AI systems, institutions and other categories and scales.

 

 

PART VI — REUSABLE DPP CASE STUDY

 

This document is also intended to become a reusable demonstration of the instrument for research, philanthropy, legal/policy, AI, organisational and commercial audiences. The Senate-submission use is one application of the specimen, not the limit of the case.

 

The reusable demonstration structure is:

 

recognisable case → raw data → conventional measurement → DPP measurement → information difference → changed target → practical consequence

 

The practical consequence is already observable. The transcript ends with an explicit churn signal: 'If not I will likely change companies.' The customer subsequently changed providers, reducing her monthly cost by approximately $30.

 

The company can measure churn after it occurs. DPP is attempting to make the structural path to that outcome visible earlier: continued cost imposed on Self; available internal moves exhausted; exit becomes the remaining Protect move; the original triangulation dissolves and a new relationship is formed elsewhere.

 

This changes the intervention target. Instead of asking only how to make the officer nicer, the script clearer, or the customer less frustrated, the organisation can ask whether its own enacted weighting is repeatedly forcing customers and staff to carry the cost of preserving a representation that does not fit the situation.

 

Same interaction. Same data. Different thing becomes measurable.

 

 

PART VII — RESEARCH STATUS / WHAT THIS CASE DOES NOT YET CLAIM

 

This is a prototype test and working case study. It does not by itself establish the universal Grammar claim, validate DPP across categories, or establish the larger isness inference. Its role is as one local measurement in a broader research programme.

 

The larger test programme can ask whether the same invariant coordinates repeatedly reveal local isness across otherwise different categories and scales; whether different Desire Signatures can be discriminated; whether mapped signatures predict later navigation or distortion before the outcome; and whether structural intervention changes the predicted behaviour.

 

In that larger machine, this deliberately boring telecommunications interaction is one cog.

​

​

How to License and Apply the deVinery Isness Parallax

​

 

What you have here is a cookie crumb, not the whole deVinery instrument. Try the crumb on your own data.

​

Take this simple Parallax test and apply it to your own material. A complaint, decision, policy dispute, AI output, meeting record, customer interaction, or any situation where different people seem to be arguing about what the problem actually is. You should be able to see something that was difficult to see before.

​

You will also hit a ceiling.

 

That is because the full deVinery Isness Parallax sits inside a larger navigation system, including the deVinery Grammar of Desire, Desire Signatures and the methods required to apply them accurately.

​

The crumb is yours to test.

 

Further use, training, commercial application and implementation require licensing through the deVinery Institute.

 

Contact the deVinery Institute to discuss licensing, training, research or collaboration.

​

​

 

​

bottom of page