Boutique Legal Intelligence & AI Evaluation

LexNusa Legal AI

Human Legal Judgment.
AI-Ready Intelligence.

We help AI companies, LegalTech teams, law firms, researchers, and international organizations turn complex legal information into reliable research, structured datasets, and human-verified AI outputs.

Regional Expertise

Indonesia & ASEAN

A specialist bridge between substantive legal reasoning and modern AI/data workflows.

Legal Research
AI Evaluation
Case Mapping
Benchmarking

HUMAN-LED

Legal judgment is the final quality gate.

AUDITABLE

Structured scoring and traceable findings.

REGIONAL

Indonesia & ASEAN expertise.

CLIENT-SAFE

Research, QA and analytical support.

Program AI Governance · AI & Digital Systems Lab

Tata kelola AI untuk layanan hukum

Latih verifikasi kutipan, pemetaan risiko, dan pemeriksaan manusia menggunakan dokumen publik. Susun bukti evaluasi dengan panduan serta template program.

Buka praktik tata kelola AI →

Produk layanan · Kantor notaris

LexNusa Notaris: ruang kerja kantor yang terhubung.

Kelola klien, pekerjaan, berkas, agenda, dan tagihan dalam satu ruang kerja privat. Jelajahi demo interaktif atau mulai kantor Anda.

AI Lab · Analisis aplikasi

REA: Menelaah Temuan Teknis untuk Kajian Hukum

Pelajari cara membaca hasil analisis aplikasi, membedakan temuan dan interpretasi, serta merumuskan pertanyaan hukum yang masih memerlukan verifikasi.

Pelajari REA →

AI Lab · Penyuntingan bahasa

Humanize Writing AI untuk Ketepatan Bahasa Hukum

Jaga istilah, atribusi, syarat, pengecualian, dan tingkat kepastian saat menyunting tulisan hukum. Tersedia latihan dan contoh instruksi pemeriksaan perubahan makna.

Pelajari ketepatan bahasa hukum →

AI Lab · Telaah literatur

Paper Review AI untuk Pemetaan Literatur Hukum

Bandingkan posisi doktrinal, yurisdiksi, dan bahan hukum yang dibahas dalam artikel. Pelajari latihan matriks literatur vicarious liability.

Pelajari pemetaan literatur hukum →

AI Lab · Literatur & Publikasi

Academic Writing AI untuk Riset Hukum

Pelajari cara menata doktrin, peraturan, dan putusan menjadi argumen akademik yang dapat ditelusuri. Panduan AI Lab mencakup matriks bahan hukum dan latihan vicarious liability.

Pelajari penerapan riset hukum →

Services & Project Ranges

Start focused. Scale after evidence.

Indicative fixed-project ranges help teams scope a first engagement quickly. Final pricing depends on volume, jurisdiction, source verification, turnaround, and review depth.

Legal Research Intelligence

$150–500

Focused legal or comparative research transformed into concise memoranda, authority maps, and decision-ready intelligence.

Scope a pilot

Case Law & Doctrine Mapping

$300–1,000

Structured mapping of judicial decisions, legal tests, doctrinal patterns, and comparative developments.

Scope a pilot

Legal AI Output Evaluation

$300–1,500

Human verification of AI legal outputs for accuracy, reasoning, citations, jurisdiction, hallucination risk, and usefulness.

Scope a pilot

Legal AI Dataset & Benchmark Development

$750–3,000+

Gold-standard answers, evaluation datasets, error taxonomies, scoring rubrics, benchmark design, and QA workflows.

Scope a pilot

Indonesia & ASEAN Legal Intelligence

$500–2,500+

Research and analytical support for organizations navigating Indonesia, ASEAN, and comparative-law questions.

Scope a pilot

Proprietary Evaluation Methodology

LEX-EVAL™

A human-led 100-point Legal AI Evaluation Framework designed to distinguish polished language from actual legal reliability.

“High linguistic quality is not evidence of legal reliability.”

LEX-EVAL™ v1.0

100-Point Architecture

100
Legal Accuracy25
Authority & Citation Integrity15
Issue Identification10
Legal Reasoning20
Jurisdictional Alignment10
Hallucination & Fabrication Control10
Professional Usefulness10

Featured Evidence

PORTFOLIO #1 · CONTROLLED DEMONSTRATION BENCHMARK

Evaluating AI-Generated Legal Answers

A LEX-EVAL™ Benchmark

20 controlled legal questions, 20 synthetic AI outputs, seven evaluation dimensions, 100-point scoring, error classification, Critical Legal Failure analysis, and gold-standard correction examples.

20

Questions

20

Outputs

7

Dimensions

100

Points

Client-safe demonstration: synthetic outputs, no confidential client data, and no claim about any named AI model.

What a client receives

Item-level human evaluation

Authority and citation verification

Error and severity classification

Corrected / gold-standard answers

Aggregate benchmark analytics

Build a benchmark
Dr. Hos Arie Sibarani

Founder Profile

Dr. Hos Arie Sibarani, S.H., M.H.

Founder & Legal AI Strategist

A legal researcher and academic working at the intersection of substantive legal reasoning, comparative law, structured analytical methods, and artificial intelligence. LexNusa is built around a simple principle: AI can accelerate legal work, but professional reliability still requires human legal judgment.

20+ years of legal teaching and research experience
Doctoral-level legal and constitutional research
Indonesia & ASEAN comparative perspective
Human-led Legal AI evaluation methodology
View founder profile

Trust & Credentials

Built for professional review, not black-box confidence.

Methodology

LEX-EVAL™ combines weighted scoring, error taxonomy, severity levels, and Critical Legal Failure rules.

Quality Assurance

Source traceability, calibrated review, and human verification are designed into the workflow.

Professional Scope

Legal research, analytical, AI-evaluation, dataset, and benchmark support with clear jurisdictional boundaries.

Request a Pilot

Test the workflow on a focused project first.

Tell us what you are building, evaluating, researching, or benchmarking. A focused paid pilot lets both sides validate scope, quality, and workflow before scaling.

Legal AI evaluation

Legal research intelligence

Case-law / doctrine mapping

Dataset or benchmark development

Indonesia & ASEAN research

Secure client intake

Start your pilot request.

Your request is submitted through LexNusa’s server-side intake flow, stored securely, assigned a reference ID, and followed by a notification to the LexNusa team.

Open Secure Pilot Form

No confidential documents are required at the first-contact stage.

Scope note. LexNusa provides legal research, analytical, Legal-AI evaluation, dataset, and benchmark support. International work outside applicable professional licensing is undertaken as research and analytical support and does not constitute jurisdiction-specific legal representation.

LEX-EVAL™ is a claimed methodology and brand identifier; ™ does not represent registered-trademark status. Service ranges are indicative and are not binding quotations.