Resources · Research archive

We thought about it
before we built it.

We have lived these problems through more than a decade in the industry. Rather than follow the AI trend, we spent more than a year working out how AI should actually solve them, and designed the answer before building HeyHR on an AI-native design philosophy. These are the papers from that work, listed with a short summary.

  • 2025research began
  • 18papers and design notes
  • 4patent applications that followed

  1. WhitepaperThe SME HR Paradox

    Why the world's largest employer segment never got a scalable HR system: fragmented inside (no policies, no specialists) and outside (rules that differ by jurisdiction). Argues the paradox can only be resolved from an AI-native foundation.

  2. Technical briefAI-native by design

    First outline of the platform: conversational access, layered architecture and security and compliance designed in from the start, not added later.

  1. WhitepaperAI-native RegTech for HR compliance

    Regulation as code. How legal text can be read, compiled and executed as machine logic, so compliance becomes continuous, verifiable and auditable across jurisdictions.

  2. WhitepaperThe post-agent era

    When anyone can build an agent, the advantage moves to domain depth and trust. Why applied AI will be won by those who understand the problem, not the tooling.

  3. WhitepaperFrom human-in-the-loop to human-out-of-the-loop

    Confidence-based autonomy. Humans stay in charge, and intervention is reduced only where evidence, telemetry and guardrails justify it.

  4. WhitepaperThe HR paradox for multi-jurisdiction SMEs

    Payroll, leave and claims as the universal entry points, and how they grow into an integrated HR operating system for companies working across borders.

  5. BriefWhy SME HR never scaled, and why it finally can

    SME HR is not a workflow problem. It is a decision problem without a reliable logic base. Software and outsourcing both inherit the same gap.

  6. ArchitectureAI capability map

    How legal data becomes executable logic: the AI layers, the deterministic platform underneath, and the foundations that connect them.

  7. Market studyWake up sleeping giants

    SME HR in Asia and Europe: vast, fragmented and still largely service-based. Why the market exists but was never captured, and what unlocks it.

  8. ArchitectureAI-native HR system architecture

    Regulation → Policy → Process → Insight. A continuous loop where AI interprets, people approve and engines execute.

  9. RoadmapBuild roadmap, 2025 to 2028

    A deliberate sequence: design first, then build, then launch in Singapore and Hong Kong, then expand country by country.

  10. WhitepaperRedefining HR for SMEs and beyond

    Not HR meets AI. AI redefines HR. The consolidated thesis: start with payroll, leave and claims, and evolve into a full HR OS fitted to SMEs.

  11. Technical noteRule compiler and multi-region compliance engine

    Making regulation as code real: rules compiled, tested and versioned, with every rule traceable to its legal source.

  12. ConceptConversational self-service for the GPT era

    Employees just want to ask. The concept that became Metis for employees: answers grounded in their own data and your policies, with privacy built in.

  13. ResearchAI-generated statutory schemas

    Testing whether modern models can define a country's payroll knowledge end to end, and what that means for how fast a payroll platform can add countries.

  1. PoC designDeterministic execution for compliance-grade AI

    Same input + same approved rule → always the same output. How approved rules are admitted, executed and reproduced, with every result attributed to its rule version.

  2. PoC designProvenance graph

    Recording how every result was produced, linking each figure to its inputs, intermediate values and governing rule version, so it can be replayed, not reconstructed.

  3. PoC designAI-native HR knowledge automation

    Using recorded calculation lineage so AI can answer "why is this number different?" from facts, not guesses.

Internal working papers. Full texts are not published. Investors and partners can ask for a briefing at ir@heyhr.ai.

See where the research ended up.

Book a demo