We build products that matter.
From automated vulnerability scanning to AI bias detection — tools for security teams, compliance officers, and builders who care about what they ship.
Automated vulnerability scanning with one-click fix-PR generation. CodeCoach layers on micro-lessons so developers learn while they ship.
Bias detection and fairness monitoring for AI systems. Compliant with EU AI Act, India DPDP Act, and GDPR out of the box.
Glass cockpit security platform for vCISOs. Multi-tenant, AI-powered insights across your entire client portfolio in one view.
Modern security scanning platform with a Go API gateway, distributed scan workers, and an MCP server for AI agent integration.
Cross-platform endpoint agent that discovers, classifies, and monitors PII and compliance-sensitive data across enterprise devices.
Enterprise policy-as-code platform with AI-powered policy generation, real-time compliance monitoring, and Kubernetes integration.
Automates evidence collection from 100+ enterprise tools and maps it to compliance controls across SOC 2, ISO 27001, GDPR, and more.
AI-powered data redaction that automatically identifies and removes PII, financial data, and trade secrets from documents and data streams.
The truth filter for the AI age. Forward suspicious content on WhatsApp and get a plain-language deepfake or misinformation verdict in seconds.
Patients send vitals and reports via WhatsApp or Telegram. ArogyaSync monitors trends and delivers AI-powered dosage and care recommendations.
AI platform for Indian classical music raga classification and generation. 95%+ accuracy across 1,616+ ragas in both Carnatic and Hindustani traditions.
Sits on top of your issue tracker and HR system to surface who is overloaded, what's at risk, and what actions will keep delivery on track.
Self-serve connector that links your Linear workspace to Power BI so teams can build custom delivery dashboards without engineering help.
Modern subscription management UI backed by hledger plain-text accounting. Your financial data stays portable, version-controllable, and yours.
AI-powered meditation app that uses Gemini to generate deeply personalized sessions with voice cloning for a truly unique mindfulness experience.
No products in this category yet — check back soon.
Independent research in AI security, privacy, governance, and applied systems.
Axonome builds empirical studies, reproducible artifacts, and working systems across AI agent security, sensitive-data discovery, secure code generation, compliance evidence, network cryptography, and AI governance.
The portfolio is intentionally claim-bounded: published artifacts are separated from active experiments, product systems are separated from paper-ready evidence, and exploratory systems stay labeled until they have reproducible benchmarks.
Portable agent skills, coding-agent rules, MCP tools, and runtime permissions create security behavior that cannot be evaluated from prompts alone.
Metadata-first and context-first discovery methods reduce raw content exposure while preserving useful recall for secrets and sensitive files.
Persistent security rules and scanner-derived guidance are evaluated for how they change downstream coding-agent behavior.
Evidence semantics, reviewer-gated promotion, AI BOM fairness evidence, and framework mapping for audit-sensitive workflows.
Network CBOMs, TLS inventory, certificate posture, and post-quantum readiness framed as longitudinal measurement problems.
Featured Work
These projects have the clearest evidence path today. Status labels stay visible so demos, products, validation packets, and published artifacts are not treated as the same thing.
Project: trust-the-skill-but-verify
Differential testing for portable AI agent skills, measuring how runtime policy changes security outcomes under the same task and contract.
Claim boundary: Method and bounded evidence, not ecosystem prevalence or full commercial-runtime coverage.
Project: llm-framing-paper
A multi-model empirical study showing that targeted security rules reduce detector-counted insecure API use in coding agents.
Claim boundary: Claims are bounded to detector-counted insecure API use and labeled validation slices.
Project: metadata-secret-scan-study
Context-first sensitive-data discovery that measures how much raw content exposure can be reduced without losing useful recall.
Claim boundary: Not a full DLP replacement. Primary privacy claims require deterministic arms and reproducible baselines.
Project: patchpilot_codecoach
A closed-loop secure-code-generation system that converts scanner findings into persistent agent rules to reduce vulnerability recurrence.
Claim boundary: Recurrence claims stay controlled until repository, agent, CWE, and functional-correctness coverage expands.
Project: grc / grclaw-standalone
Evidence semantics for AI-assisted compliance, designed to prevent workflow completion from being mistaken for audit-ready proof.
Claim boundary: Current evidence is architecture and bounded smoke evidence until reviewer or auditor evaluation exists.
Project: radar
A network CBOM and cryptographic-drift platform for measuring TLS inventory, algorithm exposure, and post-quantum readiness over time.
Claim boundary: Demo and package readiness are separate from completed external validation or longitudinal evidence.
Evidence Policy
Axonome separates claims by evidence class so security and governance work remains auditable after publication, deployment, or external review.
Experimental work, side-projects, and research explorations. Open source and built for curiosity.
AI-powered raga recommendation engine. Suggests classical music compositions based on mood, time of day, and rasa theory from ancient Indian musicology.
Open-source chess engine and AI training framework exploring reinforcement learning, neural networks, and game tree search algorithms.
Have questions about our products or want to discuss a partnership? We'd love to hear from you.