KLIQ
Computer vision platform that turns a field photo into a structured, verified record and writes it straight into Salesforce, Microsoft Dynamics or any system through a REST API. Covers retail shelves, asset inspection and nature monitoring.
What is KLIQ?
KLIQ is a computer vision platform built by KLIQ AI BV, in Antwerp, around a single promise: a photo goes in, a structured and verified record comes out, inside the business system the customer already runs.
The published pipeline has four stages. A field worker captures an image from a smartphone, tablet, drone, fixed camera or satellite tile, with nothing changing in the way they work. A vision layer then sees what is in the frame, detecting objects, labels, counts and positions. A second layer, which the vendor calls context intelligence, understands the frame against that customer's own commercial context — its SKUs, planograms, territories and field process — to decide what matters and which record it belongs to; it is configured per customer and keeps learning from that customer's corrections without human annotation. Finally the platform writes the structured fields into Salesforce, Microsoft Dynamics or any other system through the API, each output carrying its confidence score and its source photo.
Three products sit on this shared platform. ShelfSignal covers retail execution for FMCG, pharma and hospitality: share of shelf, facing counts, out-of-stock detection, planogram compliance and price tag reading. AssetScan covers inspection across six verticals — buildings, vehicles, machines, infrastructure, art and collectibles — and encodes twenty-eight established grading standards such as NEN 2767, BVRLA, ISO 10816, FHWA NBI, PSA and GIA. BioGauge scores nature assets against eighteen environmental frameworks including EUDR, FSC, VCS and TNFD.
Developers reach the platform through a REST API with official TypeScript and Python SDKs, API key authentication, webhooks and an interactive OpenAPI explorer; the documentation claims a first result in under five minutes. The Salesforce integration is detailed, writing to any standard or custom sObject with the photo attached as a File, and works alongside Agentforce and Data Cloud.
The vendor claims up to 98% detection accuracy, adds a human verification layer before results reach the customer, and positions its per-observation audit trail as built for Article 12 of the EU AI Act.
What it does
- Turn a smartphone, drone, fixed camera or satellite image into structured fields with confidence scores
- Write the result into a standard or custom Salesforce object, or into Microsoft Dynamics 365, through the API
- Count facings, read shelf prices and score planogram compliance from a single retail shelf photo
- Grade the condition of a building, vehicle, machine or piece of infrastructure against its industry standard
- Keep a time-stamped, geo-tagged audit trail with the source photo behind every data point
- Trigger alerts and action lists when compliance drops across a store network
- Score nature assets for biodiversity, forestry and carbon reporting frameworks
When to use KLIQ / When not to
A quick filter to help you decide if KLIQ is the right fit.
When to use KLIQ
- Field sales representatives and merchandisers who currently record shelf observations by hand
- Facility, maintenance and inspection teams that grade building, machine or infrastructure condition against a published standard
- Salesforce and Microsoft Dynamics teams that want field photos to land as records instead of attachments
- Developers who would rather call a vision API than build and maintain an in-house computer vision pipeline
- European buyers who need EU data residency and a per-observation audit trail for regulatory reporting
When not to use KLIQ
- Buyers who need a published API price list, since the platform tier is still described as 'pricing details at launch'
- Teams that need the native Salesforce managed package today, as it is still pending Salesforce Security Review
- Organisations whose procurement requires an achieved ISO 27001 or SOC 2 certificate, both of which the vendor lists as in progress
- Marketing Cloud and Commerce Cloud users, which the vendor states are outside the scope of its Salesforce integration
- Anyone wanting to evaluate the product alone, as no self-service demo or sandbox is reachable without contacting the team
How to use KLIQ
A typical end-to-end flow, from setup to results.
- Create a developer account on the platform, or start from the free tier of 1,000 photos per month
- Generate an API key under Settings then API Keys and store it as an environment variable
- Install the official SDK, for example npm install @kliq-ai/sdk, or the Python equivalent
- Initialise the client with the key and scope it to your tenant
- Create a location with its name, address and coordinates
- Create an observation carrying the location, the image URL, coordinates and any metadata you want to keep
- Start a computer vision job on that observation, choosing the model that fits your use case
- Read the result by polling, or subscribe to the completion webhook for production workloads
- For Salesforce, call the API from Apex, Flow or a Lightning Web Component through a Named Credential and map the fields to your objects
- For field teams, have staff photograph the shelf or asset in the app, which also captures offline and syncs on reconnection
Pros & Cons
Pros
- A genuinely documented technical offer: OpenAPI schema, two official SDKs, webhooks, guides and recipes
- The recognition model is trained on the customer's own catalogue before any contract is signed
- Every field keeps its source photo, GPS position and timestamp, which makes results auditable rather than merely stored
- EU-West data residency and a European publisher, with GDPR treated on a dedicated page
- A real free entry point with no card: 1,000 photos per month on the API, 150 starter credits on ShelfSignal
- Deep, well-documented Salesforce coverage including Consumer Goods Cloud and Field Service
- Free evaluation material: a savings estimator that asks for no email, a whitepaper and three costed scenarios
Cons
- No public price for the platform and API offer, which the site leaves at 'pricing details at launch'
- The only published grid belongs to ShelfSignal and opens on the annual view, showing 99 EUR where the commitment-free monthly rate is 119 EUR
- Neither ISO 27001 nor SOC 2 has been obtained; the vendor describes both as in progress
- The sub-processor list is not published, only available on request
- The native Salesforce package is still pending Salesforce Security Review
- No self-service demo or sandbox, and the mobile app announced for offline capture is on neither app store
- The API host named in the published OpenAPI schema does not resolve, and the site contradicts itself on whether the API has launched
Pricing & Plans
A permanent free plan exists on both sides of the offer: every account includes 1,000 photos per month on the API with no card required, and the ShelfSignal entry tier, named Scout, is free with 150 starter credits. The lowest publicly priced paid plan is ShelfSignal Business at 119 EUR per seat per month billed monthly, or 99 EUR per seat per month on an annual commitment, which the vendor presents as two months free. That tier includes 5,000 credits per seat per month and allows up to 75 active users. Enterprise is quoted individually, with an indicative floor of 149 EUR per seat per month from 200 seats. All amounts are stated in euros and exclude applicable taxes, and purchased credits do not expire. Two cautions apply. The pricing page opens on the annual toggle, so the first figure a visitor sees is the committed one. And these amounts price ShelfSignal, one of the three products: the platform and API offer carries no published figure at all beyond its free tier.
- quick scan
- one trained SKU
- personal dashboard
- mobile app with offline capture
- 5
- 000 credits per seat per month
- up to 75 active users
- full computer vision observations
- planogram compliance scoring
- REST API and CRM integrations
- SSO and SAML
- unlimited active users
- unlimited credits under fair use
- white-label option
- custom model per catalogue
- dedicated customer success manager
- EU AI Act Article 12 audit trail
- API access — free account including 1
- 000 photos per month
- with paid usage-based pricing announced but not published
Data, GDPR & hosting
A consolidated view of how KLIQ handles your data.
GDPR overview
GDPR is addressed concretely, on a dedicated page last updated in March 2026. KLIQ AI BV names itself as controller, lists the Article 6 lawful bases it relies on, and enumerates the seven data subject rights with a thirty-day response commitment at hello@kliq.photo. It commits to notifying the supervisory authority within seventy-two hours of a breach under Article 33, and names the Belgian data protection authority as the competent body. A Data Processing Agreement is offered on request to Business and Enterprise customers, and processing is said to stay within the EEA unless Standard Contractual Clauses apply. Two limits are worth noting: the sub-processor list exists but is only available on request rather than published, and no Data Protection Officer is named, the page pointing to a generic 'data protection team' at the same address.
Who owns the data?
The terms of service state that customers keep ownership of the data and images they upload, and grant KLIQ AI BV only a limited licence to process them as far as needed to deliver the service. The vendor, in turn, owns the platform itself, its features and its content. For personal data, the privacy policy names KLIQ AI BV as the data controller for everything collected through the platform; the GDPR page adds that for enterprise customers the company may instead act as a processor under a Data Processing Agreement. A single address, hello@kliq.photo, handles ownership, privacy and legal questions alike.
Reuse rights
Customers keep their uploaded images and data and may reuse them freely: the licence granted to the vendor is limited to operating the service, and nothing in the documents restricts what the customer does with its own outputs. In the other direction, the vendor states it uses account data, usage data, uploaded images and their location and time metadata, payment details held by its provider, and support messages, in order to run and improve the service, bill it, send transactional and product emails, watch for abuse and meet legal obligations. It says it does not sell personal data and shares it only with hosting, analytics and payment providers and with legal authorities. On model training the position is explicit but narrow. KLIQ trains a model on the customer's own product catalogue before a contract is signed, and its context layer keeps learning from that customer's corrections, with no human annotation. The site never says whether customer material feeds shared models, and no opt-out mechanism is documented anywhere.
Data retention & training
Hosting summary
Both the privacy policy and the GDPR page state that all data is hosted in EU-West data centres, and that processing takes place primarily within the European Economic Area. Where data does leave the EEA, the vendor says appropriate safeguards apply, specifically Standard Contractual Clauses approved by the European Commission. The commercial copy reinforces the same point, describing the product as GDPR-native with data residency in EU-West and built in Europe. Two precisions are missing. No individual country is named, so EU-West remains a region rather than a jurisdiction, and no hosting provider is identified. The sub-processor list that would answer both questions exists but is only available on request. Security measures are described in general terms: encryption in transit over TLS, encryption at rest, access controls and regular security assessments. Note that the marketing website itself is served through a content delivery network, which says nothing about where the application platform stores customer data.
Things to keep in mind
Risks and trade-offs to weigh before adopting KLIQ.
- The published pricing grid opens on the annual view, so the first figure shown assumes a twelve-month commitment; the commitment-free monthly rate is higher
- Those amounts price ShelfSignal alone: reading them as the platform or API entry ticket would misprice the offer, since the API side publishes no figure
- ISO 27001 and SOC 2 appear on the site only as work in progress, so neither should be treated as an obtained certification
- The three legal documents are titled for KLIQ Pulse and their navigation returns to a consumer section, while the privacy policy speaks of the wider platform: the contractual scope is not closed, and they are dated March 2026, before the company was registered
- The company trains a model on customer material and no opt-out is documented anywhere, so customers who need one must negotiate it
- Retailer names such as Carrefour, Delhaize, Colruyt and Albert Heijn describe the market the product targets, not a customer list
- Residual pages outside the sitemap describe stablecoin rewards and a token for public contributors, an activity unrelated to the enterprise offer presented on the homepage
Setup & Integrations
Technical difficulty
Low to moderate, and it depends on the route. For a developer the path is short: create an account, generate an API key, install the SDK, then create a location, an observation and a vision job — the documentation claims a first result in under five minutes. For a Salesforce org, expect a Named Credential and an object-and-field mapping, work an admin or integrator can do without custom code once the native package ships. The heavier part, model calibration, is handled by the vendor before signature, and the site states that no machine learning expertise, training data or GPU infrastructure is required. For field staff, capture is a single tap and works offline.
Deployment
Integrations
Supported languages
Behind KLIQ
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement KLIQ.
Frequently asked questions
What does KLIQ actually do with a photo?
Which systems does it write into?
Which Salesforce clouds are supported?
Is there a free plan?
How much does the paid entry cost?
Where is the data hosted?
How long is data kept?
Does KLIQ train on customer data?
Is KLIQ certified ISO 27001 or SOC 2?
Is there a mobile app?
Should you pick KLIQ?
KLIQ is a technically credible offer carried by a very young company. The engineering surface is real and checkable: a product-specific OpenAPI schema, official TypeScript and Python SDKs, webhooks, an interactive reference, and a Salesforce integration described down to the object and the Named Credential. The idea that separates it from generic image classification is the second layer, which carries each customer's own commercial context and keeps learning from that customer's corrections, and the audit trail that keeps the photo, the GPS position and the timestamp behind every field.
The reservations are equally concrete. KLIQ AI BV was registered on 23 April 2026, and several parts of the offer are still catching up with the marketing. The API host named in the published schema does not resolve, the homepage announces the API as live while the API page still carries a launch date and a waiting list, the native Salesforce package remains in security review, and the mobile app used in the product story appears on no app store. On assurance, the vendor is candid rather than misleading: it writes that ISO 27001 and SOC 2 are in progress, and keeps its sub-processor list available on request rather than published.
Commercially, only one of the three products carries a public grid. ShelfSignal starts free and then asks 119 EUR per seat per month, or 99 EUR with a twelve-month commitment, while the API side is left at usage-based pricing to be detailed later.
For a European team that already lives in Salesforce or Dynamics and wants field observations to arrive as records with their evidence attached, KLIQ is worth a pilot — starting on the free tier, and with the maturity gaps above written into the trial plan.
- Choosing a selection results in a full page refresh.
- Opens in a new window.