Legal Pace
Legal Pace is a Dutch legal analytics platform for law firms, combining case-outcome prediction, case law search and a private chatbot. Its models run locally inside the firm, so case files never leave the office.
What is Legal Pace?
Legal Pace, which also trades as PACE, is a legal analytics and assistance platform built in Amsterdam. Its name is an acronym the company uses throughout the site: Procedural Analytics for Case Efficiency. It sets out to answer three complaints it attributes to Dutch legal practice, namely decisions taken with little data behind them, inefficient case law research, and the hours lost summarising and structuring files. The answer is three products presented as a single chain running from assessment to research to assistance. Case Chances gives a quantitative reading of a case's success rate and its risk factors, drawing on a database the company describes as more than 200,000 historical cases. The Search Engine retrieves case law and patterns using TF-IDF, runs locally, and can be calibrated on a firm's in-house data on request. The Legal Chatbot summarises files, structures key points and drafts documents. The technical argument rests on Small Language Models: compact systems trained on legal material and optimised per practice group, which the company sets explicitly against large generic models like ChatGPT. It claims they are faster, more private, more accurate in legal context and cheaper to run, since they need no cloud infrastructure or external APIs. Training uses anonymised case law, legislation and legal literature only, and firms may optionally have models tuned on their own files and contracts. In use, a lawyer uploads documents or enters case details on a dashboard, the model compares the matter against the historical corpus, and a report comes back carrying case probability, risk analysis and estimated cost. Legal Pace claims 85% accuracy on predicting case odds while marking the figure indicative, depending on case type and data quality. Coverage today is civil, employment, rental and administrative law, with expansion promised on customer demand. Alongside the software the company sells data services, assessing a firm's data landscape, restructuring and classifying it, then building a search model over it, and positions itself as an innovation partner. The team is named openly: founders Jude Ladd and Bram van der Knaap, three specialists, and an advisory board of three.
What it does
- Estimate the odds of winning a case and the risk factors behind them
- Produce a report combining case probability, risk analysis and an estimated cost
- Search Dutch case law and surface comparable precedents in seconds
- Summarise case files and judgments automatically
- Structure the key points of a file before review
- Draft contracts and alternative clauses with AI assistance
- Tune a search model onto the firm's own document archive
When to use Legal Pace / When not to
A quick filter to help you decide if Legal Pace is the right fit.
When to use Legal Pace
- Dutch litigators weighing whether a civil, employment, rental or administrative law case is worth bringing
- In-house legal teams that need a data-backed second opinion before committing to proceedings
- Law firms bound by professional privilege that cannot send case files to a cloud-hosted general-purpose LLM
- Practices sitting on a large internal document archive they have never been able to search properly
- Paralegals and legal assistants who lose hours summarising files, judgments and precedents by hand
When not to use Legal Pace
- Lawyers practising outside the Netherlands, since the model is trained on Dutch case law and no other jurisdiction is offered
- Anyone working beyond civil, employment, rental and administrative law, the only areas covered today
- Solo practitioners looking for a self-service tool with a published price, as every plan goes through a sales quote
- Teams that need to be productive this week, given a two-to-four-week implementation and no working self-service login
- Buyers who require formal security certification, as no SOC 2, ISO 27001 or NEN 7510 audit is claimed
How to use Legal Pace
A typical end-to-end flow, from setup to results.
- Request a demonstration or a quote through the contact form, as there is no self-service sign-up
- Check scope with the team, since coverage is limited to civil, employment, rental and administrative law
- Agree a plan among Starter, Professional and Enterprise, and receive a tailored quote
- Allow two to four weeks for implementation, depending on the size of the firm
- Have the models deployed to run locally inside the firm's own environment
- Go through the personal onboarding and work from the supplied manuals
- Optionally have the search engine calibrated on the firm's in-house documents and contracts
- Upload the relevant documents, or enter the case details, on the dashboard
- Let the model compare the matter against the historical case database
- Read the returned report covering case probability, risk analysis and estimated cost
Pros & Cons
Pros
- Case data is processed locally and, the vendor states, never leaves the firm's office
- An unconditional commitment that customer case data is not used to train or improve the models
- The analysis environment is isolated and makes no external API calls
- Covers the whole chain from case assessment to research to drafting rather than one slice of it
- The search engine can be calibrated on the firm's own archive, not just public case law
- Human support is part of the offer, with personal onboarding, manuals and phone and email support
- A 30-day money-back guarantee applies to every plan
Cons
- No price is published: all three plans display Custom and route to a quote request
- There is no terms and conditions page at all, so nothing contractual governs the service
- The privacy policy is short, Dutch-only, and lacks retention periods, a subprocessor list and any procedure for exercising rights
- No data processing agreement is offered and no security certification such as SOC 2, ISO 27001 or NEN 7510 is claimed
- The product is visibly unfinished, with login not implemented, the demo slot picker marked coming soon and the demo video a placeholder
- The 85% accuracy and 200,000-case figures are self-declared, with no published method or audit
- Coverage is limited to four areas of Dutch law, and the company's stated location differs between its own pages
Pricing & Plans
Legal Pace publishes no prices. All three plans, Starter, Professional and Enterprise, are presented as Custom and lead to a quote request, and the monthly and yearly toggle on the pricing page carries no amounts behind it. There is accordingly no permanent free plan and no entry-level price point that can be stated in any currency. The commercial route is a free demonstration by appointment followed by a tailored quote. A 30-day money-back guarantee is offered on all plans.
- Starter — up to 50 analyses per month
- machine learning predictions
- basic reporting and email support
- API access and priority support are excluded. Price on quotation.
- Professional
- marked POPULAR — up to 200 analyses per month
- advanced ML models
- extended reporting
- priority support and API access
- team collaboration tools are excluded. Price on quotation.
- Enterprise — unlimited analyses
- custom ML models
- a dedicated account manager
- 24/7 support
- API access and team collaboration tools. Price on quotation.
Data, GDPR & hosting
A consolidated view of how Legal Pace handles your data.
GDPR overview
Legal Pace claims compliance in plain terms, GDPR Compliant and fully compliant with European privacy laws, and its FAQ repeats that it uses end-to-end encryption and fully complies with the regulation. Hosting is said to be on European servers and communications run over TLS/SSL. Beyond those claims the documentation is thin. The privacy page is a short privacy promise available only in Dutch, carrying no retention periods, no legal basis for processing, no procedure for exercising data subject rights, no subprocessor list and no effective date. No data protection officer is named and no Article 27 representative is designated, although a Netherlands-based controller would not need one. The single published privacy contact is info@legal-pace.com. Compliance here is asserted by the vendor, not evidenced by published documentation.
Who owns the data?
No terms and conditions are published anywhere on the site, so no contractual clause assigns ownership of the material a firm uploads. What Legal Pace states is a position rather than a warranty: case data is never stored outside an isolated analysis environment, it is neither logged nor kept once the analysis has run, and the AI executes locally inside the firm's own environment so that no legal data leaves the office. The privacy page sums this up as Your data stays private. Always. Ownership therefore appears to remain with the client, but by assertion rather than by a signed contractual term.
Reuse rights
The vendor is explicit that customer material is not recycled into its models: case data is not used to train or improve them. The models are machine learning systems, described as neural networks and tree-based models, trained solely on anonymised Dutch case law, legislation and legal literature, and no generative service such as ChatGPT is called for case input. The analysis environment makes no external API calls. One nuance matters. A firm may optionally have models tuned on its own files, contracts and internal knowledge base, which is a deliberate paid engagement and not the same thing as a general model learning from every customer. Nothing published restricts what a firm may then do with the output it obtains.
Data retention & training
Hosting summary
Legal Pace describes a two-part arrangement. The models themselves run locally, inside the client firm's own environment, and the company states that case data never leaves the office and is not kept outside an isolated analysis environment that makes no external API calls. Separately it claims European servers, described as EU servers and secure European servers, without naming a country, a data centre or a hosting provider, so the jurisdiction is stated only at regional level. Communications run over TLS/SSL and the site claims end-to-end encryption. No subprocessor list is published and no data processing agreement is offered or mentioned. As context rather than vendor claim, the marketing site itself sits behind Cloudflare and uses Supabase, which concerns the website and not the case data. For a buyer the practical position is that the sensitive processing is asserted to happen on the firm's own infrastructure, and everything beyond that is described only in broad terms.
Things to keep in mind
Risks and trade-offs to weigh before adopting Legal Pace.
- A predicted win rate is easy to mistake for advice. The 85% figure is self-declared, marked indicative, and says nothing about your specific file, so treating it as a decision rather than an input is the central risk.
- Advising a client to settle or litigate on the strength of a score invites real professional liability, and no terms and conditions page exists to define where the vendor's responsibility ends and yours begins.
- Routine reliance on automatic summaries can quietly erode the habit of reading a file closely, which is precisely the skill a lawyer is retained for.
- The privacy commitments are asserted rather than documented: no retention periods, no subprocessor list, no data processing agreement, and no certification such as SOC 2, ISO 27001 or NEN 7510.
- The company's stated location differs across its own pages, Amsterdam in the footer, contact page and structured data but Leiden on the privacy page, and no legal form is named anywhere.
- The privacy page claims no analytics tracking cookies while the consent banner announces cookies used to analyse services, and the two statements do not agree.
- The client testimonials use conspicuously generic Dutch personal and firm names, and the trusted-by banner names no firm, so the social proof cannot be verified.
Setup & Integrations
Technical difficulty
Moderate, and not self-service. Legal Pace quotes two to four weeks for implementation depending on the size of the firm and its requirements, and the models are deployed to run locally inside the firm's own environment, which implies involvement from whoever manages that infrastructure. There is no online sign-up, since the login page states the function is not yet implemented, so access begins with a sales conversation. The company offsets this with personal onboarding, manuals and ongoing support by phone and email, and its data services are delivered as guided engagements rather than something a firm configures alone.
Deployment
Supported languages
Behind Legal Pace
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Legal Pace.
Frequently asked questions
How secure is my data with Legal Pace?
How accurate are the predictions?
Which areas of law does it cover?
Is my case data used to train the models?
What does it cost?
Can I try it before deciding?
How long does implementation take?
Is there an API?
Should you pick Legal Pace?
Legal Pace answers a real constraint. A Dutch lawyer who wants machine help on a file cannot simply paste it into a general-purpose chatbot, and this platform is built around that fact: the models run locally inside the firm, the analysis environment makes no external calls, and the company commits without qualification that customer case data is not used to train or improve anything. For a practice bound by professional privilege that architecture is the point, and it is more than most legal AI vendors offer. The product itself is coherent. Predicting case odds, searching case law and drafting from a private model are three things a litigator genuinely does in sequence, and covering all three rather than one is a sensible design. The optional calibration on a firm's own archive is the kind of feature that separates a tool from a demonstration. The reservations concern maturity and documentation rather than intent. Nothing is priced publicly, so no reader can judge cost without a sales conversation. There is no terms and conditions page at all, and the privacy policy is a short Dutch-only promise with no retention periods, no subprocessor list and no route for exercising data subject rights, which is thin ground for a product handling privileged material. The 85% accuracy figure is self-declared and marked indicative. The site still shows its seams: login is not implemented, the demo slot picker says coming soon, and the client testimonials carry conspicuously generic names. The sensible reading is a young and credible Dutch legal tech company whose privacy engineering runs ahead of its paperwork. If you practise civil, employment, rental or administrative law in the Netherlands, the free demonstration costs nothing. Ask for the contractual terms, the retention rules and the basis for the accuracy figure before committing.
- Choosing a selection results in a full page refresh.
- Opens in a new window.