High-stakes preparation
Bar preparation is resource-intensive and emotionally demanding, especially for refreshers and graduating students who need repeated feedback.
AI can help Bar takers. It must also be checked.
AI Bar Helper is a verification-first coaching and readiness system for Philippine Bar preparation. It helps students structure answers, identify missing elements, and verify legal claims before relying on them.
Built from MSc FinTech dissertation research and responsible AI readiness work. Designed to support, not replace, faculty teaching and human judgment.
For incubator stakeholders: this MVP is designed to show three things quickly - a clear customer pain, a differentiated verification-first workflow, and a practical path from service-led pilots to a scalable institutional product.
01
The problem is not only answer generation.
Bar preparation is high-stakes, AI use is already happening, and the weakest link is often verification discipline.
AI without citation habits
Students increasingly use AI tools, but many workflows do not build citation checking, source awareness, or revision discipline.
Confident but incomplete
Frontier models can sound persuasive even when a rule is incomplete, unsupported, outdated, or simply wrong.
Schools need readiness
Law schools need workflow design, policy guidance, and verification habits, not just another chatbot.
02
A coaching workflow, not an answer shortcut.
AI Bar Helper helps learners structure, check, and improve their bar-style answers. It does not replace faculty, reviewers, or primary legal sources.
Student works from a Bar-style question or enters a draft answer.
The system checks whether the answer follows ALAC: Answer, Legal basis, Application, Conclusion.
Feedback identifies missing issues, weak reasoning, unclear phrasing, and revision opportunities.
Every legal claim is treated as unverified unless supported by authorized source text.
The student improves the answer and owns the final work.
The product is not the answer. The product is verification.
03
Try the Coach My Answer demo
Choose a subject, answer a short Bar-style practice question, and receive coaching on structure, missing points, and verification flags.
Practice Question
Student Answer
Coach My Answer
Feedback
For study support only. Not legal advice.
Verify all authorities before relying on them.
Do not paste confidential client information.
Only paste source materials you are authorized to use.
AI outputs may be incomplete or wrong.
Feedback for improvement
Submit a practice answer to receive coaching feedback.
How scoring works
Loading rubric...
04
RAG is a hypothesis, not the assumption.
The MVP is designed to test whether frontier models, structured prompting, pasted-source verification, RAG, or hybrid workflows give the best balance of cost, safety, traceability, and usefulness.
| Approach | Strength | Tradeoff |
|---|---|---|
| Frontier model only | Cheap and fast | Weaker traceability |
| Prompted ALAC mode | Cheap and structured | Still needs verification |
| Pasted-source mode | Source-aware and safer for early pilots | Manual source preparation |
| RAG mode | Traceable and institution-controlled | Higher setup and maintenance cost |
| Hybrid verification mode | Separate source-checking layer reviews claims | Best tested after pilot evidence |
05
Service-led pilot packages
The first commercial form is consulting, training, and pilot implementation for law schools and bar review groups.
AI Readiness Talk
60 to 90 minute session for students or faculty on responsible use, hallucination risk, citation checking, and verification habits.
Verification-First Bar Workshop
Guided answer coaching session covering ALAC structure, source-checking, revision drills, and AI-use boundaries.
Pilot Design Sprint
School-specific AI use policy notes, prompt checklist, verification checklist, diagnostics template, and pilot metrics planning.
06
Built for pilot learning, not premature scale.
AI Bar Helper is an early-stage, service-led LegalTech and EdTech venture with a practical path from pilots to scalable institutional software.
Beachhead customer
Philippine law schools, bar review groups, faculty mentors, and refresher cohorts that need responsible AI readiness and answer-coaching workflows.
Initial wedge
Paid talks, workshops, and pilot design sprints that teach verification-first AI use while generating product-discovery data.
What the pilot will test
Whether students improve structure, identify missing legal elements, and build better verification habits; and whether faculty find the workflow safe and useful.
Early business model
Service-led pilots first, followed by institutional workshops, consulting, and later subscription or licensed school deployment after validation.
Metrics to track
Coached answers, revision completion, missing elements by subject, unverified claims flagged, satisfaction, and pilot conversion inquiries.
Incubator support needed
Pilot design mentorship, pricing guidance, IP and data governance review, architecture validation, and legal education partnerships.
07
Evidence and responsible-use basis
Built by Leo Ernesto Thomas G. Romero, CPA, lawyer, and MSc FinTech graduate, from dissertation research on the feasibility of an AI Bar Helper for Philippine Bar preparation.
Conservative research basis
The concept was dissertation-tested across comparative answers, student usability, and practitioner feedback. The key insight is simple: AI outputs can be useful and persuasive, which is exactly why verification matters.
Product direction should test frontier-only, prompt-only, pasted-source, RAG, and hybrid verification workflows before deciding on expensive architecture.
Pending and exploratory institutional work may inform future pilots, subject to approvals and institutional policies.
Privacy notice
The demo processes the question, answer, and optional source text to generate coaching feedback. The MVP does not store answer submissions by default.
Users should not paste confidential client data or sensitive personal information. Contact form data is used only to respond to pilot inquiries.
08
Request a pilot conversation
For law schools, faculty mentors, review coordinators, refreshers, and institutional stakeholders exploring responsible AI readiness.