One integration.
Every counterparty.
hirinq is built on open standards so your ATS, talent app, or AI agent can participate without proprietary lock-in.
A2A — Agent-to-Agent
hirinq implements Google's Agent-to-Agent (A2A) protocol, enabling hiring agents to communicate across platforms. An employer agent posts a structured job; a talent agent evaluates it against a candidate's profile. Both sides negotiate within defined constraints, with humans reviewing final decisions. No proprietary message format. No platform lock-in. Just the standard.
Standard message envelope
Every negotiation round uses the A2A task envelope — typed, versioned, and machine-readable by any A2A-compatible agent.
Task-based negotiation
Each hiring negotiation is a discrete A2A task with defined states: pending, active, matched, rejected. Both sides advance state independently.
Cross-vendor interoperability
Employer agents on Workday, talent agents on a boutique platform — A2A means they negotiate without a shared backend.
MCP — Model Context Protocol
hirinq exposes its hiring queue, candidate pipeline, and negotiation state as MCP tools. Any MCP-compatible AI assistant — Claude, GPT, Gemini via MCP client — can read pipeline status, propose shortlists, and trigger negotiation rounds, all through Anthropic's standard Model Context Protocol interface. Your assistant speaks to hirinq the same way it speaks to any other MCP server.
Read/write access to queues
Create, inspect, and update hiring queues from any MCP client. Your assistant can report pipeline status and surface blockers without leaving its context.
Candidate evaluation tools
Score an OTP profile against an OJP spec, retrieve the rationale, and surface shortlists — all as native MCP tool calls with typed inputs and outputs.
Negotiation triggers
Advance negotiation state, request candidate consent upgrades, and accept matches — all without leaving the assistant conversation.
BYOM — Bring Your Own Model
Don't want to use hirinq's default matching model? Swap it for your own. Connect any model that can score an OTP profile against an OJP spec. Your model returns a score and rationale; hirinq handles the bilateral clearing, audit trail, and GDPR compliance. Your model stays in your infrastructure — hirinq never sees your weights.
Standard scoring interface
OTP + OJP in, score + rationale out. The interface is minimal and stable. If your model can HTTP, it can integrate.
EU AI Act Article 12 audit trail
Regardless of which model scores the match, hirinq logs every decision, prompt, constraint check, and state transition. Compliance is infrastructure-level, not model-level.
Your model stays in your infra
hirinq calls your endpoint. Your model weights never leave your environment. You control latency, data residency, and version rollout.
OTP — Open Talent Protocol
An open, MIT-licensed schema for representing a candidate as structured, portable data. Skills, experience, availability, consent, and disclosure tier — all in a standard format any platform can adopt without permission. hirinq uses OTP as the canonical talent-side data format for every match and negotiation.
Portable by design
OTP profiles are self-contained JSON documents. A candidate can carry their profile across any platform that supports the protocol — no export, no re-entry.
Disclosure tiers
Three levels of data exposure — metadata (anonymous), profile (structured professional data), contact (identifiable) — governed by candidate consent, not platform policy.
MIT licensed
No vendor lock-in. Any ATS, talent app, or AI agent can implement OTP without a commercial agreement. Spec and reference implementation at opentalentprotocol.org.
OJP — Open Job Protocol
The job-side counterpart to OTP. A structured spec for a role: required skills, seniority, location, compensation band, and hiring constraints — in a format designed for machine-readable bilateral matching. hirinq uses OJP for every hiring queue. Your ATS pushes a spec; hirinq matches it against talent-side OTP profiles deterministically.
Deterministic matching
OJP's structured constraints — must-have skills, experience years, salary band, location — produce repeatable scores against any OTP profile. Same spec, same result, every time.
GDPR-compliant processing
OJP encodes the legal basis and data minimisation requirements for each role. hirinq enforces them at match time — only the disclosure tier the role requires is ever accessed.
MIT licensed
Open spec, no gating. Any employer, ATS, or job board can publish OJP-compliant postings. Spec at openjobprotocol.org.
Works with the tools you already use.
hirinq connects to your existing ATS. Structured job specs flow in via the Open Job Protocol (OJP). Candidate data stays in your system — hirinq reads what it needs via OTP.
Don't see your ATS? Contact us — we'll scope the integration.