Price intelligence for rural collection points in South Africa
Know today's price before the truck leaves.
Weighbridges and volume sensors at collection points, plus price reports sent from ordinary phones, feed one trusted daily picture of what produce is worth and where demand is. PriceGrid drafts the market advisory and the value-chain analysis. A cooperative manager approves every advisory before it goes out.
Load-cell scale1,240 kg
Volume sensor86 crates
Phone price reportR 9.80 / kg median
Quality cameraGrade A 72%
Illustrative example. Values on the cards are not live telemetry.
WeighScales and volume sensors record what arrives.
ReportTraders send prices from any phone.
AdviseA daily advisory, in plain words.
ApproveA person signs off before it is sent.
Original illustration: what arrives is weighed, the price is posted, and the same message reaches people who can't be there.Original illustration: with both prices in view, a trader can decide where to take the load. The prices shown are examples.
Who it is for
Cooperatives and collection points
A fair daily price picture and a one-click advisory for members.
Processors and retail buyers
Supply volumes and quality grades from the source, before trucks are booked.
District agricultural offices
A record of what smallholders faced each week, for planning and support.
Funders and programmes
Measured change in prices received, with an audit trail of who approved what.
This solution requires current AI models to operate. PriceGrid's advisories, forecasts, value-chain analysis and photo grading are produced by AI models, with a person approving every output. The demo on this site is simulated. The production system would not work without those models, or their current equivalents.
What PriceGrid does
Combines weighed volumes and trader price reports into an anonymous daily picture.
Drafts a plain market advisory and a value-chain analysis for a person to review.
Grades produce from photos and flags disagreements for checking.
What it does not do
It does not set or fix prices, or tell anyone what to charge.
It does not show any trader's own price by name.
It does not guarantee a price or give financial advice.
It does not handle payments or contracts.
The problem
Without a price signal, everyone sells blind.
A smallholder or informal trader often learns the price only when a buyer names it. When many people go to the same place with the same produce, prices drop and good produce spoils, and nobody keeps a record to learn from.
Produce is readyHarvest will not wait.
No price signalOnly the buyer knows.
Sell to whoever comesLittle room to negotiate.
Everyone goes one waySame produce, same market.
Local glutPrice drops fast.
Produce spoilsOr is sold below cost.
No recordHard to plan next time.
Three situations the system tells apart
Normal swing
Prices move a little from day to day within their usual range. The advisory says to carry on.
Local glut
Volumes jump and the price falls at one point while another market holds up. The advisory lays out options for the manager to choose from.
Not enough reports
Too few reports or a disagreeing pair. The system holds back the price and asks for more, rather than guessing.
Estimates, not verdicts. PriceGrid gives people an estimate to decide from. It does not set prices, promise outcomes or replace a trader's own judgement.
How it works
From a weighbridge reading to an advisory a manager signs off.
Ingest
Weighbridges and volume sensors record what arrives. Traders and farmers send prices by WhatsApp, SMS or a simple app.
Aggregate
Reports are cleaned and combined into a median price for each product and point. Outliers are flagged and a price is shown only when enough reports support it.
Explain
A language model drafts the daily advisory in short, plain sentences, with the evidence behind it.
Pattern-check
Seasonal history and the value chain show whether a price move is normal and where value is lost between farm and shelf.
Verify
A vision model grades produce from photos, and weighed volumes are cross-checked against reported supply.
Visualise
A price board, a dashboard and a short phone message show the same facts in the format each person uses.
A person approves
The cooperative manager approves, modifies, rejects, asks for more reports, or escalates. Nothing is sent before that.
Fair data, fair prices
Aggregates, not individuals.Only combined prices are shown. No trader's own price is displayed by name.
Enough reports first.A price appears only when enough independent reports support it. The minimum is set per commodity during the pilot.
See what a price is based on.Every price shows how many reports and weighings stand behind it.
Built to inform, not to fix.South Africa's Competition Commission has guidelines on sharing competitively sensitive information between competitors, and they favour aggregated and historical data. The design is reviewed against that guidance before launch.
What we would install
Device
What it measures
Power
Connection
Design notes
Load-cell scale
Weight of each delivery
Solar panel with battery
Short-range radio to the gateway
Zeroed at opening, checked against a known weight
Volume sensor
Crates or bags counted at the gate or bay
Battery
Short-range radio
For points without a scale
Price reporting
Prices reported by traders and farmers
Their own phones
WhatsApp, SMS or app
Works on basic phones through SMS
Quality camera
Photos of sample crates for grading
Solar panel with battery
Cellular or phone upload
Fixed viewpoint and lighting for repeat photos
Field gateway
Collects readings and stores them if the signal drops
Solar panel with battery
Cellular, satellite where there is none
Locked, weatherproof and raised
Rural coverage is patchy, so every device keeps working offline and sends later. Hardware models, suppliers and maintenance are agreed with each pilot site.
The six roles
Five helpers, one person who decides.
Each helper does one job, so each can be checked on its own. None sends a message, sets a price or signs anything. Here is what each reads, produces, how it helps, and what a person still checks.
1. Price aggregator
Statistical model
Reads
Price reports from traders and farmers, weighbridge volumes, the time and the product grade.
Produces
A median price per product and point, the number of reports behind it, and flags for outliers and disagreements.
How it helps
Replaces a scatter of rumours with one number people can trust, and shows how firm it is. It holds back a price when reports are too few.
A person checks
Flagged outliers, which may be an error or a real local bargain, and reports that look coordinated.
2. Supply and price forecaster
Forecasting model
Reads
Weighed volumes, past prices, season, weather and the number of reporting points.
Produces
A short outlook for supply and price at each point, with a confidence value, and an early warning when a glut is forming.
How it helps
Gives managers a day or two of warning to spread loads across markets before prices fall.
A person checks
Local events the model cannot see, such as a funeral, a strike, a road closure or a festival.
3. Advisory writer
Language model
Reads
The aggregated prices, the forecast, and the options the manager has used before.
Produces
A daily market advisory in short, plain sentences, ready to send by WhatsApp or SMS, in the languages members use.
How it helps
Saves the manager an hour each morning, keeps messages clear and consistent, and shows the evidence in one line.
A person checks
Tone, wording, language and whether the advice is fair to every member before it is sent.
4. Value-chain analyst
Long-context language model
Reads
Months of prices at the farm gate, the collection point and the market, transport costs and spoilage records.
Produces
A value-chain analysis: where margin is earned and lost, which routes pay best after transport, and what grading or storage could change.
How it helps
Turns price data into a case for action that a cooperative can bring to buyers, funders or the district office.
A person checks
Whether the costs and margins match what members know, and whether the conclusions are fair to buyers as well.
5. Produce quality checker
Vision model
Reads
Photos of sample crates taken at the same spot and in similar light.
Produces
A grade estimate, such as size, colour, bruising and ripeness, with a confidence value. It also flags a poor photo.
How it helps
Gives farmers a neutral second opinion on quality, so grade disputes start from evidence, and helps sell better grades to buyers who pay for them.
A person checks
Doubtful grades, which an operator confirms by hand. Light, shade and camera angle can all change what the model sees.
6. The cooperative manager
Human reviewer (mandatory)
Reads
All of the above, plus what no sensor can see: who depends on which market, past arrangements and local knowledge.
Decides
Approve, modify, reject, ask for more reports, or escalate to the district agricultural officer.
How it helps
Keeps responsibility with a named person who members know and can question.
Advised by
The market committee and the district agricultural officer, where there is one.
Model candidates for each role
Candidates under evaluation, not endorsements and not purchases. Family names are used because specific versions change often, and each is checked against the provider's current documentation before use. Nothing on this site calls these models: all AI output in the demo is simulated.
Role
OpenAI / Azure
Google Cloud
AWS
NVIDIA-related
Price aggregator
Azure Machine Learning
Vertex AI, BigQuery ML
Amazon SageMaker
RAPIDS on GPU instances, if volume needs it
Supply and price forecaster
Azure Machine Learning forecasting
Vertex AI Forecast
SageMaker time-series models
NVIDIA GPU instances for training
Advisory writer
GPT family through Azure OpenAI
Gemini through Vertex AI
Amazon Bedrock (Amazon Nova, Anthropic Claude and others)
NVIDIA NIM microservices for self-hosted open models
Value-chain analyst
Long-context GPT models
Long-context Gemini models
Long-context models on Bedrock
NVIDIA NIM
Produce quality checker
Vision-capable GPT models; OpenAI image models for synthetic training photos
Gemini vision; Imagen for synthetic training photos
NVIDIA TAO Toolkit and Jetson for edge grading at the point
Data ingest
Azure IoT Hub
Google Cloud IoT-compatible partners and Pub/Sub
AWS IoT Core
NVIDIA Jetson gateways, optional
Planning cost range (estimate)
A planning estimate for a pilot of about eight collection points, in rand per month, before any grant or programme credits. It is not a quote. It will be replaced by figures from each provider's pricing calculator once the pilot design is fixed.
Item
Indicative range per month
What drives it
Cloud ingest, storage and dashboards
R 2,000 to R 8,000
Number of devices and how often they report
Advisory and analysis language models
R 1,000 to R 6,000
Advisories per day, languages, and length of history analysed
Photo grading
R 500 to R 4,000
Photos per day and whether grading runs at the edge
Messaging (WhatsApp and SMS)
R 1,000 to R 5,000
Members reached per advisory
Connectivity
R 1,000 to R 4,000
Cellular data plans, satellite where needed
This solution requires current AI models to operate; the models listed (or their current equivalents) are essential to the production system. Model identifiers are checked against current provider documentation. Outputs shown on this site are simulated.
A worked example (simulated)
An illustration of how the roles fit together. Not real data.
Readings arrive
A collection point's scale shows tomato volume 2.4 times higher than last week, and 22 traders report lower prices.
Aggregator finds the price
The median price is 28% lower than last week, supported by 22 reports, with two outliers flagged.
Forecaster warns of a glut
Volume is expected to stay high for two more days at this point, while a nearby market is quiet.
Advisory writer drafts three options
Hold half the load for a day, take a load to the nearby market, or sell graded produce to a processor.
The manager decides
The manager approves the second option with changes, and the message goes to members.
The value-chain analyst reports later
A month on, the analysis shows how much the rerouting saved after transport costs.
What the models never do
Send a message or move a load without a person's approval.
Set, fix or recommend a price for any individual.
Show one trader's report to another by name.
Present an estimate as a guarantee.
Hide uncertainty. Every output shows confidence and what it cannot see.
How we keep them honest
Every price shows how many reports and weighings stand behind it.
Operators confirm doubtful photo grades by hand.
Named model families below are candidates under evaluation, not endorsements. Exact versions are checked against provider documentation at build time.
Deployment options under consideration: Azure IoT Hub with Azure OpenAI, AWS IoT with Amazon Bedrock, or Google Cloud with Vertex AI, with optional edge processing at the collection point.
3D model
Pile on the produce and watch the price move.
Drag to look around, scroll or use the buttons to zoom, and move the slider to change how much produce arrives at the collection point.
Scale sensor
Price board
Price message
Collection point
Market B
The 3D model needs WebGL and an internet connection to load its library. Try another browser or check your connection.
Weighed load4.0 t
Price on the boardR 11.00 / kg
StatusSTABLE
Proposed truck actionWaits at the point
Manager boardNo action needed
Flat platform with glowing corners: the scale and its load cells.
Orange crates: produce being weighed.
Tall board: today's price for everyone to see.
Phone on the stand: the price message people receive.
Spinning wireframe ball: the price model.
Dark board at the back: the cooperative manager, who decides.
Yellow truck: shown as a proposal, not a dispatch.
Stall on the right: Market B.
How this model works, in plain English
This is a small toy version of one collection point. The slider decides how much produce arrives, and the scale counts it crate by crate. As the pile grows past what the market can absorb, the price on the board drops and the status changes from stable to attention to high-risk. The wireframe ball stands for the price model, which notices the change and drafts an advisory. The dark board at the back is the cooperative manager, who sees the advisory and decides. The yellow truck only drives to Market B as a proposal, because nothing moves until a person approves. Switch to Assembly view to pull the parts apart and see what each one does.
Simplified for teaching. The price shown is an example, not a real price. Real prices depend on the product, season, quality and demand. In real use the manager approves before any load moves.
Impact
Why this matters.
The price a farmer gets is often decided by who knows more on the day. PriceGrid is built to narrow that gap without taking the decision away from anyone.
KM
A note from the founder
I started KTJ AI Solutions because farmers and small traders in rural South Africa often sell without knowing what their produce is really worth a few kilometres away. I want every cooperative and collection point to see a fair, trusted price before they load the truck. We are building PriceGrid so that information reaches people on the ground, in plain language, even where the signal is weak.
Kitso Tshepo Joseph Mofokeng, Director, KTJ AI SOLUTIONS
What it is designed to change for each person
Farmers and traders
A fair, plain-language price before the truck leaves, on any phone, including basic ones through SMS. Neutral photo grades to start grade disputes from evidence.
Cooperative managers
An hour back each morning, a clear view of where a glut is forming, and a record of who approved what.
District agricultural offices
A weekly picture of what smallholders in the district faced, to target support where prices fall hardest.
Funders and buyers
An audit trail of approvals and a measured before-and-after, and for buyers, supply and quality signals from the source.
Why we think it can work
Cooperative managers already decide, every day, where loads go and what to tell members. PriceGrid does not replace that job. It gives the manager better evidence sooner and a drafted message, then waits for approval. Because the human-in-the-loop step matches existing practice, the pilot needs a new tool, not a new way of working.
Honest scope. These are design intentions for a prototype, not measured results. There is no production data behind them yet. The research on the Research page also warns that knowing a price does not by itself raise incomes where buyers set prices, so PriceGrid is one tool and not a cure. Measured impact will be reported only from real pilot data.
Research and evidence
Information gaps are documented. So are the limits.
In one study of smallholder irrigation schemes, 59% of independent irrigators could reach both informal and formal markets, while 30% of public-scheme irrigators could reach only informal markets.
From a 2021 paper on small-scale farming and market access in South Africa. The same literature warns that knowing the price does not by itself raise incomes, especially where buyers set prices and transaction costs are high, so PriceGrid is one tool and not a cure.
How each part of PriceGrid helps
Scales and volume sensors give a neutral record of supply. Price reports from phones spread the price signal. Quality grading and the value-chain analysis give producers evidence to bargain with. The approval step keeps a named person responsible for what is sent.
A legal explainer on South African guidelines that shape how price information can be shared between competitors. A reason PriceGrid shows aggregates, not individual prices.
Articles
Ten problems, and how PriceGrid would help.
Each article states a real problem for rural markets, how the AI and the Solution Bot would help, and which model families could do the work. They are educational notes for a prototype, not press releases.
01Why a smallholder often learns the price only when the buyer names it
Problem
A farmer or informal trader arrives at a collection point with produce and no independent price to compare. The buyer sets the number, and the seller has little room to negotiate. The same produce can sell for very different prices a few kilometres apart on the same day.
How this AI + Bot solution improves it
PriceGrid combines weighbridge readings and trader price reports into one median price per product and point, and shows how many reports stand behind it. The Solution Bot answers "what is the price and how firm is it?" in plain words, and it holds the price back when reports are too few. A manager approves the message before members see it.
AI models that could improve this
GPT / OpenAI via Azure OpenAIGoogle Gemini on Vertex AIAmazon SageMakerAWS Bedrock
Candidates under evaluation, not endorsements.
02Seeing a local glut forming one or two days before prices fall
Problem
When many people take the same produce to the same place, supply outruns demand, the price drops and good produce is sold below cost. By the time the price board shows it, the loads are already there.
How this AI + Bot solution improves it
The supply and price forecaster watches weighed volumes against history and raises an early warning when a glut is forming. The advisory writer drafts options, such as holding half the load or taking a load to a nearby market. The manager approves, modifies or rejects, and the bot explains the next steps in the order to follow.
AI models that could improve this
Amazon SageMaker forecastingVertex AI ForecastGPT / OpenAI via Azure OpenAINVIDIA NIM / TAO / Jetson
Candidates under evaluation, not endorsements.
03Grade disputes at the gate: giving both sides a neutral second opinion
Problem
Whether a crate is grade A or grade B decides the price, and the grader and the farmer often disagree. Without a shared record, the dispute is one person's word against another's.
How this AI + Bot solution improves it
A quality camera photographs sample crates from a fixed viewpoint. A vision model estimates size, colour, bruising and ripeness, with a confidence value, and flags poor photos. Doubtful grades go to an operator for a hand check. The bot tells the operator how to take a usable photo and what to do when the model is unsure.
AI models that could improve this
Google Gemini on Vertex AIGPT / OpenAI via Azure OpenAIAmazon Rekognition Custom LabelsNVIDIA NIM / TAO / JetsonImage models (OpenAI, Imagen)
Candidates under evaluation, not endorsements.
04Heat, rain and spoilage: timing loads around the weather
Problem
Fresh produce spoils faster in heat, and rain slows trucks and raises transport costs. Sellers decide on the day without a joined-up view of weather and price.
How this AI + Bot solution improves it
PriceGrid already reads live public weather for candidate collection areas. In production, weather would feed the forecaster, and the advisory would flag hot afternoons for fast-spoiling produce such as spinach. The bot suggests which loads to sell early and which to hold, always as options for a manager to approve.
AI models that could improve this
Amazon SageMakerVertex AIGPT / OpenAI via Azure OpenAIAWS Bedrock
Candidates under evaluation, not endorsements.
05Price reporting from any phone, in the languages members use
Problem
Many traders have basic phones, patchy signal and costly airtime. An app-only tool leaves them out, and a message in a language they do not use is not a message at all.
How this AI + Bot solution improves it
Prices come in by WhatsApp, SMS or a simple app. Language models read short free-text reports and turn them into structured prices for review, and draft advisories in the languages members use. The bot walks a new trader through how to send a report and what happens to it.
AI models that could improve this
GPT / OpenAI via Azure OpenAIGoogle Gemini on Vertex AIAmazon Translate and BedrockNVIDIA NIM for self-hosted open models
Candidates under evaluation, not endorsements.
06Sharing prices without breaking competition rules
Problem
Competitors sharing detailed price information can raise competition-law concerns. South Africa's Competition Commission has published guidelines on exchanging competitively sensitive information, and they favour aggregated and historical data.
How this AI + Bot solution improves it
PriceGrid shows only anonymous aggregates and never displays a trader's own price by name. A price appears only when enough independent reports support it. The bot explains these rules in plain language and points to the guidance, and the design is reviewed against it before launch. It does not replace legal advice.
AI models that could improve this
GPT / OpenAI via Azure OpenAIGoogle Gemini on Vertex AIAWS Bedrock Guardrails
Candidates under evaluation, not endorsements.
07Keeping weighbridge numbers honest
Problem
A scale that drifts, or a count that does not match what was reported, quietly distorts every price built on top of it.
How this AI + Bot solution improves it
Scales are zeroed at opening and checked against a known weight. Weighed volumes are cross-checked against reported supply, and anomalies are flagged to the operator. The bot gives the operator a daily checklist and tells them what to do when a scale is offline.
AI models that could improve this
Amazon SageMaker anomaly detectionAWS IoT CoreVertex AINVIDIA NIM / TAO / Jetson
Candidates under evaluation, not endorsements.
08Where value is lost between farm gate and shelf
Problem
A cooperative may know that margins are thin, but not whether the loss is in transport, spoilage, grading or the final sale. Without evidence it is hard to bargain or to ask for support.
How this AI + Bot solution improves it
The value-chain analyst reads months of prices at the farm gate, the collection point and the market, with transport costs and spoilage records. It produces a draft analysis of where margin is earned and lost. A person checks the costs against what members know before it is used with buyers or funders.
AI models that could improve this
GPT / OpenAI via Azure OpenAIGoogle Gemini on Vertex AIAWS Bedrock long-context modelsNVIDIA NIM / TAO / Jetson
Candidates under evaluation, not endorsements.
09Working where the signal is weak: offline-first devices
Problem
Rural coverage is patchy. A tool that stops when the network drops is not a tool for the places that need it most.
How this AI + Bot solution improves it
Every device keeps recording offline and sends later through a locked, weatherproof field gateway. Where there is no cellular signal, satellite is an option. The bot tells operators what to check first when a device goes quiet, and what the manager can still do in the meantime.
AI models that could improve this
AWS IoT Core and IoT GreengrassAzure IoT Hub and IoT EdgeNVIDIA NIM / TAO / JetsonVertex AI edge models
Candidates under evaluation, not endorsements.
10Evidence for district offices and funders: who approved what
Problem
Programmes need to know what smallholders faced each week and whether a tool changed anything. Without a record, it is hard to learn or to justify support.
How this AI + Bot solution improves it
Every advisory keeps a record of the evidence, the model's stated confidence and limits, and the named person who approved, modified or rejected it. The analyst drafts a summary for the district office. The bot explains how pilots are measured. Impact is reported only from real pilot data.
AI models that could improve this
GPT / OpenAI via Azure OpenAIGoogle Gemini on Vertex AIAWS BedrockAmazon QuickSight
Candidates under evaluation, not endorsements.
This solution requires current AI models to operate. Model families named here are candidates under evaluation. Final choices follow provider documentation and a pilot.
Coverage
Weather moves produce and roads.
Current temperature and the next three days of rain for six candidate collection areas, because heat speeds spoilage and rain slows trucks. This is the only live data on the site. It comes from Open-Meteo and refreshes every 10 minutes while this page is open. If your browser or host blocks outside requests, the page says so instead of showing numbers.
Loading live weather…
The "Prospective" stage is the company's own designation, not derived from the weather feed.
About and team
Who is behind PriceGrid.
PriceGrid, a product of KTJ AI SOLUTIONS. KTJ AI SOLUTIONS, a company registered in South Africa. Its details below are taken from the SARS Notice of Registration, which is included further down this page.
I started KTJ AI Solutions because farmers and small traders in rural South Africa often sell without knowing what their produce is really worth a few kilometres away. I want every cooperative and collection point to see a fair, trusted price before they load the truck. We are building PriceGrid so that information reaches people on the ground, in plain language, even where the signal is weak.
Funding and investment
PriceGrid is a product of KTJ AI SOLUTIONS. The startup currently has no external funding and requires investment to begin the new project and move from this working prototype into production. Microsoft Azure for Startups, AWS Activate / AWS Startups, NVIDIA Inception, and peer AI and cloud startup programmes are the natural partners for this stage because PriceGrid is built on the same stack those programmes exist to accelerate: modern AI models for scoring, advisory generation, long-context pattern analysis and vision verification, plus cloud infrastructure for secure data ingest and human-in-the-loop decision support. Grant or programme support from these partners directly funds the next phase — real backend, real integrations, and pilot deployments with municipalities or enterprise buyers — while the human-always-decides architecture already demonstrated on this site keeps governance and liability clear. Without that support the project remains a high-fidelity prototype; with it, KTJ AI SOLUTIONS can deliver a production system that depends on current AI models and that those programmes are designed to help scale.
Business documents
SARS Income Tax: Notice of Registration
KTJ AI SOLUTIONS · Registration 2026/694274/07 · Tax reference 9822798196 · Issued 8 September 2026
The official registration notice issued by the South African Revenue Service, available as a PDF.
Contact
Talk to KTJ AI SOLUTIONS.
Pilots, tenders, partnerships or a personal question: choose the form that fits and send it. Your email program opens with the message ready to send.
Thank you.
Your email program should now be open with your message ready. Press send there to complete it. If nothing opened, write to info@ktjaisolutions.co.za directly.
This site is operated by KTJ AI SOLUTIONS, a private company registered in South Africa. Registration number 2026/694274/07. Registered office: 1234 Anahita Avenue, Centurion, Gauteng, 0157. Contact: info@ktjaisolutions.co.za.
What this website is
PriceGrid is a prototype and demonstration. Everything inside the demo app is simulated: prices, volumes, predictions, alerts and AI outputs. The only live data is the weather on the Coverage page, fetched from Open-Meteo, and it is labelled with its source and time. Nothing here is a price list, a guarantee or financial advice.
Personal information the Contact form collects
Business form: organisation, name, email, optional phone, message.
Personal form: name, email, optional phone, message.
We use it only to reply to your enquiry.
How the form works and how long data is kept
The form does not send anything to a server. When you submit, your own email program opens with the message filled in, and the message travels only if you press send there. This site stores nothing. Once an email reaches us, we keep it only as long as needed to deal with your enquiry and for ordinary business records, then delete it.
Cookies and local storage
This site sets no cookies and does not use local storage. The demo's state lives in your browser's memory and disappears when you close or reload the page. The Solution Bot runs entirely inside the page: your questions are not sent anywhere.
These services receive your IP address and browser details when your browser loads their content, as with any website that uses them.
Your rights
You may ask us to confirm what personal information we hold about you, to correct it, or to delete it. Write to admin@ktjaisolutions.co.za and we will respond. If you are unhappy with how we handle your information, you may complain to the Information Regulator of South Africa.
Jurisdiction
KTJ AI SOLUTIONS is a South African company and processes personal information in line with the Protection of Personal Information Act, 2013 (POPIA). South African law applies to this site.
AI and demo disclaimer
Predictions and AI outputs in the demo are simulated, are not professional advice, and must not be relied on for any decision. In PriceGrid's design a person always decides: nothing is sent, approved or moved without a human choosing to do it.
DEMO MODE — ALL TELEMETRY, PREDICTIONS AND AI OUTPUTS ON THIS SCREEN ARE SIMULATED
Sign in to the demo
Demo access only. No account is created or verified.
