Lasso
Lasso is an AI-native product information platform for ecommerce teams. It imports messy supplier files, enriches attributes and descriptions, generates and cleans product images, and monitors competitor catalogues, all from one collaborative editor.
What is Lasso?
Lasso is a product information management platform built around AI agents rather than retrofitted with them. It is published by DoneByAI s.r.o. under the BANDITS brand, from Prague, and aims at ecommerce teams whose catalogues have outgrown spreadsheets and manual data entry.
The platform centres on a spreadsheet-style online editor where a catalogue lives across configurable product schemas, views and attribute types. Familiar Excel behaviour is preserved — copy, paste, keyboard shortcuts — and the AI writes deterministic formulas on request. On top of that grid, Lasso runs what it describes as one AI agent per cell: right-click a value to have it filled or corrected, enrich a whole column, or fill several columns from a single call. Bulk work runs thousands of agents in parallel.
Data arrives from any supplier source. CSV, Excel, XML, JSON and PDF catalogues are ingested while the AI recognises patterns, maps columns onto the customer's schema and extracts attributes, including from unstructured files. Missing information is then recovered from the web and from supplier sites through a scraping layer that keeps a reasoning log and an audit trail on every result, which is how the vendor supports its hallucination-free claim.
Beyond attributes, Lasso writes SEO copy in the customer's brand voice, upscales, crops, cleans and re-backgrounds product photography, and can generate a product image outright. A translation glossary and localisation workflows extend a catalogue into new markets. Competitive features cover product discovery, price comparison, stock monitoring, dynamic repricing with margin protection and catalogue gap analysis, while a GEO score measures how well each product is formatted for generative search engines as well as classic SEO.
Everything is reachable programmatically. A REST API, still in beta, exposes four endpoints; an npm SDK covers TypeScript and Python; and a plug-and-play MCP server lets any AI agent request product data in real time. Real-time collaboration with roles, permissions and versioning, single sign-on, audit logs and configurable zero data retention complete the platform. The vendor advertises up to 10,000 attributes per schema, more than 50,000 products enriched per day and 99.8% uptime.
What it does
- Import any supplier file — CSV, Excel, XML, JSON or PDF — and map it automatically to your product schema
- Fill missing attributes and rewrite descriptions in bulk with parallel AI agents
- Generate SEO copy in your brand voice and translate a catalogue for a new market
- Create, upscale, crop, clean and re-background product images
- Categorise products and validate them against your own taxonomy
- Scrape the web to recover missing data, with a full audit trail on every value
- Monitor competitor prices, stock and catalogues, and reprice product by product
When to use Lasso / When not to
A quick filter to help you decide if Lasso is the right fit.
When to use Lasso
- Ecommerce and catalogue managers who onboard thousands of SKUs from supplier files every month
- Buyers, merchandisers and category planners who need consistent attributes across an entire assortment
- Marketplace operators who must categorise, tag and price user-submitted or reseller listings automatically
- Content and SEO teams producing product copy at volume in a defined brand voice
- Developers and data engineers building enriched product datasets, recommendation engines or AI shopping assistants through an API or an MCP server
When not to use Lasso
- Merchants with a handful of products, for whom a credit-based enrichment platform is far heavier than the job requires
- Teams looking for a storefront, an order management system or a CRM: Lasso only handles product data
- Anyone who needs a mobile app, since no iOS or Android application is offered
- Organisations that must contractually pin down where their data is stored, as no hosting country or region is disclosed
- Buyers who need a signed uptime commitment on an entry-level plan, the SLA being reserved for Enterprise customers
How to use Lasso
A typical end-to-end flow, from setup to results.
- Create an account from the 'Start for free' button; no payment card is requested
- Connect a source or import a supplier file — CSV, Excel, XML, JSON or PDF
- Let the AI recognise the file structure, map its columns and extract the attributes it finds
- Define your product schemas, content rules, brand voice and translation glossary
- Enrich cell by cell with a right-click, by column, by multi-column row, or in bulk through an automation
- Generate or enhance product images: creation from scratch, upscaling, cropping, cleaning, realistic backgrounds
- Review the AI reasoning log and audit trail, and correct anything by hand in the editor — manual edits cost no credits
- Invite colleagues, assign roles and permissions, and work on the same catalogue simultaneously
- Publish through a native platform integration, or export by CSV or API
- For developers: generate an API key, install the SDK for TypeScript or Python, or plug the MCP server into your own agent
Pros & Cons
Pros
- Covers the entire product data chain — import, structuring, enrichment, imagery, translation, publishing and competitive monitoring — instead of one slice of it
- Permanent free plan with 200 credits a month and no payment card required
- Pricing is published operation by operation, with a cost-per-product calculator rather than a 'contact us' wall
- Traceability is designed in: AI reasoning log, audit trail on web enrichment, validation against your own taxonomy
- Clear data posture: no training on customer data, zero-retention option, SOC 2 Type II claimed, AES-256 at rest and TLS 1.2+ in transit
- Works for non-technical catalogue teams and for developers alike, through a spreadsheet UI on one side and an API, SDK and MCP server on the other
- European vendor under Czech law, with unlimited team members from the Pro plan and in-person workshops on offer
Cons
- Credit-based billing makes the real cost depend on volume — 5 credits per imported file, 2 per enriched cell, 15 per generated image — so budgets have to be modelled rather than read off a price list
- Extra credits reset at the end of each billing period and do not roll over
- The €69 headline price is the annual-billing rate; the monthly rate is not published as static content
- No hosting country or region is disclosed, and the privacy policy provides for transfers outside the EEA
- No named subprocessor list is published, only generic categories of service providers
- No mobile app, and the API is still in beta with a waitlist on newer features
- The uptime figure shown on the site (99.8%) differs from the 99.9% SLA, which is contractual only for Enterprise customers
Pricing & Plans
A permanent free plan is available at €0, including 200 credits per month, and no payment card is required to start. The cheapest paid entry point is the Pro plan at €69.00 per user per month when billed annually, which includes 5,000 credits per month; additional credit packs begin at €70 per month on the same annual terms. Enterprise pricing is quoted on request.
- Free — €0
- 200 credits per month
- data import
- structuring and enrichment
- online editor
- parallel AI agents
- web enrichment with audit trail
- SEO copy and image enhancement
- Pro — €69 per user per month billed annually
- 5
- 000 credits per month
- higher-throughput processing
- native platform integrations
- priority support
- unlimited team members
- Enterprise — custom pricing quoted annually
- custom enrichment workflows
- account manager and in-person workshops
- compliance and security audits
- advanced roles and permissions
- custom development
- + 5 000 credits — €70 per month
- billed annually
- + 10 000 credits — €120 per month
- billed annually
- + 25 000 credits — €180 per month
- billed annually
- 50 000+ credits — custom pricing on request
Data, GDPR & hosting
A consolidated view of how Lasso handles your data.
GDPR overview
GDPR coverage is explicit and reasonably detailed. The privacy policy, effective 21 January 2026, states that processing follows Regulation (EU) 2016/679 and applicable Czech data protection law, and lists four legal bases: contract performance, legitimate interests, consent and legal obligation. The full set of data subject rights is enumerated — access, rectification, erasure, restriction, portability, objection and withdrawal of consent — with privacy@banditshq.com as the contact point and a stated 30-day response time. The competent supervisory authority is named as the Czech Office for Personal Data Protection. International transfers outside the EEA are covered by Standard Contractual Clauses. A Data Processing Agreement is offered on request to enterprise customers. No Article 27 representative is designated, and none is required: the controller is established inside the European Union. Compliance with the EU AI Act is claimed alongside GDPR.
Who owns the data?
Customers keep full ownership of what they upload. The terms state that users retain all ownership rights in their Content and that DoneByAI claims no ownership of it, while the Output generated from that Content also belongs to the customer. In exchange, the vendor receives a limited, non-exclusive, worldwide licence to process, store and transmit the Content for the sole purpose of delivering the service; that licence ends when the Content is deleted or the account is closed. Content is treated as confidential and is not accessed, used or disclosed beyond what service delivery, applicable law or the terms permit. The declared data controller is DoneByAI s.r.o., company number 23610361.
Reuse rights
Customers may reuse their own Content and the enriched Output freely, without asking permission, since both remain their property: the enriched catalogue can be published, exported by CSV or API, and pushed to any sales channel. One restriction applies — Output may not be used to train competing AI models without the vendor's express written consent. On the vendor's side, personal data serves service delivery and improvement, account and billing management, service and marketing communications, usage analytics, security and legal compliance. Personal data is not sold. It is shared only with service providers for hosting, payment processing and analytics under contractual safeguards, with partners where the customer consents, with authorities when legally required, and in the event of a business transfer. Customer data is not used to train the vendor's models unless the customer explicitly asks for it.
Data retention & training
Hosting summary
The site does not disclose where customer data is hosted. No hosting country, region or cloud provider is named anywhere in the privacy policy, the terms or the product pages. What is stated is that data may be transferred to and processed in countries outside the European Economic Area, with Standard Contractual Clauses approved by the European Commission or other legally recognised mechanisms applied to those transfers. Hosting providers appear only as a generic category of service providers bound by contractual data protection obligations. On the security side, the vendor documents AES-256 encryption at rest, TLS 1.2+ in transit, regular security audits and penetration testing, access controls and authentication measures, and a SOC 2 Type II certification. Jurisdictionally, the controller is DoneByAI s.r.o., established in the Czech Republic, subject to Czech law and to the GDPR, with the Czech Office for Personal Data Protection named as supervisory authority. The public IP of the website resolves to an Amazon CDN node, which says nothing about where customer catalogues are stored.
Things to keep in mind
Risks and trade-offs to weigh before adopting Lasso.
- Where your data physically sits is not documented: no hosting country or region is named, and the privacy policy explicitly allows transfers outside the EEA under Standard Contractual Clauses
- No named subprocessor list is published, only the generic categories of hosting, payment processing and analytics, which makes a third-party risk review difficult to complete
- SOC 2 Type II certification is asserted in the privacy policy but no trust page, report or attestation is available to check it against
- The official contact addresses sit on banditshq.com, the parent brand's domain, rather than on productlasso.com, and no support email address is published at all
- Budget can drift without anyone noticing: credits are consumed per operation, are shared across the whole organisation, and expire at the end of each billing period without rolling over
- Trusting enriched attributes wholesale is risky — the vendor's own terms warn that AI output may contain errors and place the burden of reviewing it on the customer, so the audit trail is there to be used, not admired
- Output may not be used to train competing AI models without written permission, a restriction worth noting by anyone planning to build their own product data model
Setup & Integrations
Technical difficulty
Low for the intended audience. Signing up takes no payment card, and the platform is deliberately aimed at catalogue and operations staff rather than engineers: the interface behaves like a spreadsheet, and the vendor states that no technical resources are needed and that it handles integrations and setup itself. Small teams are said to go live almost immediately, complex setups within two weeks with workshops. A more technical route exists but is optional — API keys, an SDK for TypeScript or Python, and an MCP server.
Deployment
Integrations
Supported languages
Behind Lasso
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Lasso.
Frequently asked questions
Is the free plan really free, or is it a trial?
How does the credit system work?
Will my catalogue be used to train the vendor's AI models?
Who owns the enriched product data Lasso produces?
Do I need a technical team to get started?
Which ecommerce platforms does Lasso connect to?
Is there an API, and what state is it in?
Where is customer data hosted, and what is the minimum age to use the service?
Should you pick Lasso?
Lasso addresses a problem most ecommerce teams recognise instantly: supplier files arrive in every shape imaginable, attributes are missing, descriptions have to be written, images need cleaning, and someone ends up doing all of it by hand in a spreadsheet. The answer proposed here is a product information platform where AI agents work directly inside a familiar grid, from a single cell up to a whole catalogue, with a reasoning log and audit trail attached to every value they touch. That combination — spreadsheet ergonomics for catalogue teams, an API, SDK and MCP server for engineers — is what sets it apart from both general-purpose chatbots and established PIM platforms.
The commercial terms are unusually transparent for this category. A permanent free plan, a published per-operation credit table and a cost-per-product calculator let a buyer model the spend before talking to anyone. The data posture is equally clear: no training on customer data, a zero-retention option, customer ownership of both input and output, SOC 2 Type II claimed, and a European vendor operating under Czech law.
The reservations are those of a young company. DoneByAI was incorporated in 2025 and raised a €400,000 pre-seed round from Miton; the API is still in beta; no hosting country or region is disclosed and transfers outside the EEA are provided for; no named subprocessor list exists; and the SOC 2 claim is asserted in the privacy policy without a trust page or report behind it. Budgeting also requires attention, since credits expire each period and the headline €69 price assumes annual billing.
For a catalogue that has genuinely outgrown manual work, Lasso is worth a serious evaluation, and the free plan makes that evaluation cheap. For a few dozen product records, it is more machinery than the job needs.
- Choosing a selection results in a full page refresh.
- Opens in a new window.