01 / LEARN
Legal AI Academy
Partners, associates and support staff learn source-grounded research, verification, matter hygiene, safe prompting and escalation.
View curriculumCaldr Legal AI / Train + Build + Operate
Make the firm's knowledge queryable.
Train your people to use AI safely, then turn approved precedents, playbooks and internal know-how into a source-grounded knowledge environment your team can interrogate with citations.
• Source-grounded answers• Human review gates• Client-controlled deployment
What is our preferred limitation of liability position for mid-market SaaS agreements?
The approved commercial playbook favours an aggregate liability cap tied to fees paid in the preceding 12 months, with specified carve-outs requiring partner review.
Commercial Playbook §7.2SaaS Precedent v4 · clause 11
THE MODEL
Learn it. Build it. Operate it.
The academy is the adoption layer. The implementation is the high-value work: information architecture, access boundaries, source curation, testing and rollout inside the firm's own approved Google environment.
02 / BUILD
Knowledge Desk Setup
We map practice areas, approved sources, roles and permissions, then configure the firm's knowledge environment and operating rules.
See implementation03 / OPERATE
Ongoing AI Desk
Source updates, usage reviews, prompt libraries, new workflow design, refresher training and governance improvements.
See operating modelPRACTICE AREA EXAMPLES
Useful on day one.
The desk is structured around the way lawyers already think: practice area, matter type, approved precedent, source and reviewer.
CommercialClause positions, SaaS/MSA precedents, liability caps, negotiation playbooks.
PropertyLease precedents, reports on title, CPSE material, drafting notes and checklists.
EmploymentSettlement agreements, policy wording, approved letters and internal guidance.
LitigationChronologies, witness inconsistencies, pleadings, authorities and evidence navigation.
CONTROL FIRST
The sensitive material stays out of the academy.
Firebase Hosting delivers the training and Caldr interface, with Google sign-in and Firestore for academy progress. Client documents belong in the law firm's own approved Google Workspace / Google Cloud environment. Enterprise deployments can use Google Cloud identity, IAM and region controls, subject to the firm's own information-security and professional-obligations review.
Firm-owned environmentKnowledge sources are configured within the client's selected Workspace / Google Cloud architecture rather than uploaded to Caldr's public training site.
Role-led accessPractice area and notebook access can be designed around role, team and matter boundaries instead of one giant shared corpus.
Grounded outputsUsers are trained to verify answers against cited source material and escalate when authority or version status is uncertain.
Enterprise optionFor higher-control environments, Gemini Notebook Enterprise can be deployed within the firm's Google Cloud project with IAM and enterprise controls.
No silent automationThe first release is a research and drafting desk with explicit human approval gates — not an autonomous legal decision-maker.
Governance before rolloutInformation classification, retention, sharing, privilege and client obligations are reviewed before production source ingestion.
Founder-led where judgement matters.
Nick's voice opens the highest-value modules — why this matters commercially, what good looks like, where firms get AI wrong and when to stop. Procedural walkthroughs stay concise and screen-led.
NICK'S VOICE · SELECTED MODULES
COMMERCIAL OFFER
Training is the front door. Implementation is the product.
A firm can start with the academy, then move into a scoped knowledge-desk implementation with ongoing support.
Legal AI Academy
- Partner and staff foundation training
- Six practical modules
- Source-grounded research exercises
- Verification and escalation framework
- Completion evidence and implementation readiness score
Knowledge Desk Implementation
- Discovery and information map
- Practice-area source architecture
- Identity / role design
- Prompt and query library
- Regression testing and governance pack
- Staff launch and ongoing optimisation
From scattered know-how to an AI knowledge desk.
Start with a single practice area, prove the workflow, then expand. Caldr handles the training, implementation pattern and operating discipline.