Research Paper

Why It Matters

Many enterprises rely on tacit expert rules for auditing and compliance, which are hard to document and improve systematically. Trace2Policy automates rule refinement to boost decision accuracy and reduce costly expert hours, enabling scalable, transparent, and reliable compliance processes. This transforms manual, error-prone workflows into efficient, self-evolving systems.

Potential Customers & Pain Points

• Logistics companies – Need accurate scalable audit decision-making
• Legal firms – Require consistent compliance rule application
• Enterprises in regulated industries – Struggle with costly manual compliance reviews
• Software vendors – Seek to integrate explainable decision automation.

Research Paper

Why It Matters

Organizations face increasing regulatory and privacy constraints on data use that current query correctness tools do not address. Data Flow Control ensures data safety policies are enforced directly within database queries, preventing unauthorized data combinations and releases. This approach scales across multiple database systems with minimal performance impact, transforming data governance workflows and reducing compliance risks.

Potential Customers & Pain Points

• Enterprises with sensitive data – Risk of regulatory non-compliance
• Cloud data platform providers – Need scalable data governance
• AI service providers – Ensuring privacy in automated data analysis
• Financial institutions – Preventing unauthorized data sharing
• Healthcare organizations – Protecting patient data privacy.

Research Paper

Why It Matters

Public legal departments face staff shortages and rising case volumes, causing delays and inconsistent legal advice. LegalCheck streamlines drafting processes, enabling faster turnaround and consistent application of laws, which improves operational efficiency and legal compliance at scale in government workflows.

Potential Customers & Pain Points

• Municipal legal departments – Overwhelmed by case volume and staff shortages
• Public-sector legal teams – Need faster consistent legal document drafting
• Government agencies – Require compliance with complex regulations under time pressure

Research Paper

Why It Matters

Organizations deploying LLMs in regulated sectors face legal and reputational risks from policy breaches. This solution offers a lightweight, scalable way to detect nuanced policy violations without costly retraining or latency, improving oversight and compliance workflows. It enables enterprises to enforce internal policies reliably, reducing risk and operational burden.

Potential Customers & Pain Points

• Enterprises in legal finance healthcare – Need reliable low-latency policy compliance monitoring
• AI platform providers – Require scalable interpretable violation detection
• Regulators – Demand transparent AI governance tools.

Research Paper

Why It Matters

Enterprises deploying LLMs in regulated and mission-critical environments require robust, real-time safety systems that handle diverse data types and provide explainability. Protect enhances trust and compliance by preventing harmful outputs across modalities, enabling scalable and auditable AI workflows in sensitive domains.

Potential Customers & Pain Points

• Enterprises – Need multi-modal safety and compliance for LLMs
• Regulated industries – Require explainable and auditable guardrails
• AI platform providers – Need scalable production-ready safety solutions
• Security teams – Need real-time detection of prompt injection and data leaks.

Research Paper

Why It Matters

Legal professionals rely on accurate document retrieval to support case analysis and decision-making. Current RAG systems often retrieve incorrect documents due to structural similarity in legal texts, causing errors and inefficiencies. This solution scales to large legal datasets, improving trust and workflow efficiency in legal AI applications.

Potential Customers & Pain Points

• Law firms–Inaccurate document retrieval slows case preparation
• Legal tech companies–Need reliable AI tools for legal research
• Corporate legal departments–Require precise information retrieval to reduce risk
• Courts and public legal services–Demand trustworthy AI assistance for document handling.

Research Paper

Core Innovation

This paper introduces the A2AJ Canadian Legal Data project as a fully open-source alternative to CanLII, overcoming its restrictions on bulk and programmatic access. It provides extensive legal datasets and APIs enabling new computational law applications and democratizing access to legal information. This approach supports both judicial processes and legal technology development focused on underserved populations.

Potential Customers & Pain Points

• Legal Tech Developers Needing Open Data Access
• Courts Requiring Evidence-Based Assessment Tools
• Legal Aid Organizations Serving Low-Income Communities

Research Paper

Core Innovation

This paper introduces HISPASpoof, the first large-scale dataset focused on Spanish synthetic speech detection and attribution. It addresses the gap where existing detectors trained on English fail on Spanish. The dataset includes diverse Spanish accents and multiple zero-shot TTS systems, enabling improved detection and method attribution.

Potential Customers & Pain Points

• Security Agencies Needing Synthetic Speech Detection In Spanish
• Media Companies Verifying Authenticity Of Spanish Audio Content
• AI Developers Lacking Spanish Synthetic Speech Benchmarks

Research Paper

Core Innovation

This paper introduces a hybrid architecture combining CNNs, LSTMs, and Bahdanau attention with contrastive learning to enhance feature discrimination for MPAA rating prediction. It achieves high accuracy in distinguishing fine-grained rating differences, especially for borderline cases. The approach improves generalization over prior single-model or non-attentive methods.

Potential Customers & Pain Points

• Streaming Platforms Needing Automated Content Rating
• Content Moderation Teams Seeking Consistent MPAA Classification
• Regulatory Bodies Requiring Accurate Rating Enforcement

Research Paper

Core Innovation

This paper fine-tunes a BERT model with a recall-oriented approach to improve multi-label narrative classification in news articles. It integrates a GPT-4o pipeline to enhance prediction consistency and introduces a ReACT framework with semantic retrieval-based few-shot prompting for grounded narrative explanations. The use of a structured taxonomy table as auxiliary knowledge uniquely improves classification accuracy and explanation reliability.

Potential Customers & Pain Points

• Media Analysts Needing Accurate Narrative Detection
• Educational Institutions Seeking Narrative Understanding Tools
• Intelligence Agencies Requiring Reliable Narrative Explanations

Research Paper

Core Innovation

This paper introduces TRUST-VL, a unified vision-language model that integrates a Question-Aware Visual Amplifier to extract task-specific visual features for misinformation detection. It is trained on TRUST-Instruct, a large dataset with structured reasoning chains that mimic human fact-checking workflows. This approach enables strong generalization across multiple distortion types and provides explainable outputs, unlike prior models focused on single distortion types.

Potential Customers & Pain Points

• Fact-Checking Organizations Needing Efficient Multimodal Verification
• Media Companies Combating Fake News
• Social Media Platforms Reducing Misinformation Spread
• News Aggregators Seeking Automated Content Validation
• AI Developers Improving Misinformation Detection Models

Research Paper

Core Innovation

This paper introduces GUARD, a framework that translates abstract AI ethics guidelines into concrete test queries. It uniquely uses adaptive role-play and jailbreak scenarios to uncover both direct and indirect guideline violations. This approach improves detection accuracy and provides detailed compliance reports beyond prior static testing methods.

Potential Customers & Pain Points

• AI Developers Needing Compliance Testing
• Regulators Monitoring AI Ethics
• Enterprises Deploying Safe AI Systems

Research Paper

Core Innovation

This paper introduces a context-aware query translator integrated into a RAG pipeline specifically designed for legal research. It leverages open-source SBERT and GTE embeddings to reduce costs while maintaining high retrieval quality. The approach improves answer faithfulness and reproducibility compared to existing proprietary systems.

Potential Customers & Pain Points

• Law Firms Needing Accurate Legal Research
• Legal Departments Seeking Cost-Effective Tools
• Legal Tech Companies Developing AI Solutions
• Academic Researchers in Legal Informatics
• Paralegals Requiring Faster Document Retrieval

Research Paper

Core Innovation

This paper introduces a multi-agent AI system that integrates writing, legal verification, and formatting agents to automate SOW creation. It uniquely combines retrieval-augmented generation with specialized agents to reduce drafting time from days to minutes while enhancing legal compliance. This approach surpasses prior single-agent or manual methods by improving speed and accuracy simultaneously.

Potential Customers & Pain Points

• Legal Departments Needing Faster Contract Drafting
• Business Teams Struggling with SOW Accuracy
• Consulting Firms Requiring Compliance-Checked Documents

Research Paper

Core Innovation

This paper presents NyayaRAG, a framework combining retrieval-augmented generation with legal knowledge specific to Indian common law. It uniquely integrates factual case descriptions, statutes, and semantically retrieved prior cases to simulate courtroom reasoning. This improves both prediction accuracy and explanation quality compared to prior models.

Potential Customers & Pain Points

• Law firms needing faster case analysis
• Legal tech companies seeking Indian law solutions
• Courts requiring decision support
• Legal researchers needing structured case retrieval

Research Paper

Core Innovation

This paper introduces LLM-based text embedders that handle longer inputs and work unsupervised for prior case retrieval. It overcomes limitations of traditional IR methods and supervised transformer models by enabling more effective and scalable retrieval. The approach is validated on multiple benchmark datasets, showing superior performance.

Potential Customers & Pain Points

• Law firms needing faster case research
• Legal tech companies seeking improved retrieval accuracy
• Courts requiring efficient prior case referencing
• Legal researchers lacking scalable unsupervised retrieval tools

Research Paper

Core Innovation

This paper introduces Compliance Brain Assistant, which dynamically switches between fast context retrieval and a full agentic mode to handle complex compliance tasks. This dual-mode approach improves accuracy and response quality compared to baseline large language models while maintaining low latency. It uniquely integrates composite actions and tool invocations within a conversational AI framework for enterprise compliance.

Potential Customers & Pain Points

• Enterprises with complex compliance workflows needing faster task completion
• Compliance officers requiring accurate context-aware assistance
• Legal teams managing regulatory queries under time constraints

Research Paper

Core Innovation

This paper introduces a novel RAG system that integrates large language models with hybrid search and relevance boosting to improve query accuracy and efficiency. It demonstrates superior performance over traditional RAG approaches on expert-annotated queries. The paper also offers practical hyperparameter tuning guidance for real-world applications.

Potential Customers & Pain Points

• Regulated Industry Compliance Teams Facing Complex Regulations
• Quality Assurance Departments Overwhelmed by High Query Volumes
• Risk Management Professionals Needing Accurate Regulatory Insights

Research Paper

Core Innovation

This paper presents a novel method combining textual entailment and in-context learning to translate legal text into a canonical, executable Python format. Unlike prior work, it captures both structural and semantic metadata and their relationships, enabling automated compliance verification. The approach is validated on multiple U.S. state laws with high accuracy.

Potential Customers & Pain Points

• Small and Medium Software Companies Lacking Legal Expertise Needing Compliance Automation
• Legal Tech Firms Seeking Advanced NLP Tools for Law Interpretation
• Compliance Officers Struggling with Manual Regulation Translation