Describe
Tell us what data you need in plain language. Our planner agent turns it into a typed schema and a source map.
Now onboarding design partners · Amsterdam
Web data, collected by AI agents.
Describe the data you need. Our agents find the sources, check they may be used, extract and validate every record — and keep it fresh. No personal data, ever.
agent run · eu-prices-daily
› plan "daily prices of e-bikes across NL/DE/BE"
◆ planner: 14 candidate sources · schema Product@v1
✓ compliance: robots.txt ok · ToS ok · 0 PII fields
✓ extract: 12,480 records · 3 quarantined
↻ layout change on source #2 → re-learned
✓ delivered → warehouse.prices ▌
Product demo · sample run
- 0
- personal data fields collected
- 100%
- public, licensed or permitted sources
- 3
- nationalities, one Amsterdam team
- 24/7
- monitored agent pipelines
✕Public catalogues✕Open registries✕Tender portals✕Licensed APIs✕Open data portals✕Price feeds✕Public filings✕Product docs✕Public catalogues✕Open registries✕Tender portals✕Licensed APIs✕Open data portals✕Price feeds✕Public filings✕Product docs
How it works
From a sentence to a living dataset in four steps.
Verify
A compliance agent checks robots.txt, terms of use and licences, and filters out anything that could be personal data.
Collect
Extraction agents navigate, parse and adapt when layouts change — no brittle CSS selectors to babysit.
Deliver
Clean, deduplicated, validated records land in your API, warehouse, S3 bucket or spreadsheet on schedule.
Platform
Scrapers break. Agents adapt.
Self-healing extraction
When a site changes its markup, agents re-learn the structure and keep the feed alive instead of failing silently.
Schema-first output
Every record is validated against your schema. Bad rows are quarantined with a reason, never mixed into your data.
Compliance built in
Polite crawling, rate limits, source allow-lists and full audit logs for every request we make.
Change monitoring
Get alerted when prices, catalogues, listings or documents change — with a human-readable diff.
Any destination
REST API, webhooks, Postgres, BigQuery, Snowflake, S3, Google Sheets or plain CSV.
EU infrastructure
Processing and storage in EU data centres. No personal data, no surprises for your DPO.
Use cases
What teams build with agentic data.
Example pipelines we design for. Want yours to be our first published case study?Join the design partner program.
E-commerce01
Competitor price & assortment tracking
Daily snapshots of public product catalogues across EU marketplaces, normalised into one SKU-level feed.
output → SKU · price · stock · promo flags
Real estate02
Commercial listings intelligence
Aggregate public commercial property listings, deduplicate across portals and track time-on-market.
output → listing · m² · €/m² · status history
Public sector & ESG03
Tenders and regulatory filings
Monitor public procurement portals and registries, classify notices by sector and alert on new matches.
output → tender · CPV code · deadline · value
Pricing
Start small. Scale to enterprise.
Starter
For teams validating a data idea.
€490/month
- ✓Up to 3 sources
- ✓Daily refresh
- ✓CSV / Sheets / API delivery
- ✓Email support
Growth
For products that run on fresh data.
€1,900/month
- ✓Up to 25 sources
- ✓Hourly refresh
- ✓Webhooks & warehouse sync
- ✓Change alerts
- ✓Shared Slack channel
Enterprise
Dedicated pipelines with SLAs.
Custom
- ✓Unlimited sources
- ✓Custom SLAs & uptime
- ✓On-prem / VPC option
- ✓DPA & security review
- ✓Dedicated engineer
All plans include a 14-day pilot. Prices exclude VAT (21% BTW).
Team
Four founders, three countries, one Amsterdam team.
MS
Max Schepard
Co-founder & CEO
Product, partnerships and agent orchestration.
TD
Tomáš Dvořák
Co-founder & CTO
Prague → Amsterdam. Distributed crawling infrastructure and LLM systems.
JN
Jakub Nowicki
Co-founder, Engineering
Kraków → Amsterdam. Extraction agents, data quality and validation.
PZ
Piotr Zieliński
Co-founder, Growth
Warsaw → Amsterdam. Customers, compliance and go-to-market in the EU.
FAQ
Compliance first. Questions welcome.
Is this web scraping legal?+
We only collect publicly available, non-personal data from sources whose terms permit it, or data you have a licence to. Every source passes a compliance check before any collection starts.
Do you collect personal data?+
No. Our pipelines are designed to exclude personal data entirely, and records that look like personal data are dropped automatically.
How fast can we start?+
A typical pilot with 1–3 sources is live within a week of agreeing on the schema.
What happens when a website changes?+
Our agents detect structural changes, adapt the extraction and flag anything that needs human review — you keep getting data.
Let’s build your data pipeline.
Tell us what you need. We reply within one business day with a source map and a pilot proposal.
max-schepard@amstelparsers.eu