How Archer uses AI

A zero-integration, low-risk way for Enterprises to adopt AI in recruitment

Executive Summary

Archer enables organisations to benefit from artificial intelligence in recruitment without integrating into internal systems, accessing internal candidate data, or automating hiring decisions.

Archer operates externally to your technology environment. Live roles are made available via a structured feed, typically XML, which Archer consumes in order to understand open positions. There is no direct integration into your ATS and no access to internal candidate records.

Candidates independently choose to apply and enter your ATS through your standard application process. From that point onwards, your recruitment governance, assessment criteria and decision-making processes apply in full.

Archer enhances how candidates are identified and presented with relevant opportunities. It does not alter how hiring decisions are made, and it does not constitute automated decision-making as defined under GDPR Article 22.

How Archer Works

Archer operates outside of your internal systems and interacts only with:

• A structured role feed you expose externally, typically via XML

• The hackajob candidate community

Step 1 - Role Feed

You expose your live roles through a structured feed, most commonly XML. Archer consumes this feed to understand vacancy information.

Archer does not:

• Integrate into your ATS

• Access candidate records held within your systems

• Retrieve interview notes, screening outcomes or assessment data

• Write data back into your systems

The data flow is strictly limited to reading vacancy information that you choose to publish via the feed.

Step 2 - Role Calibration

Archer uses artificial intelligence to analyse each role across four dimensions:

• Job-fit requirements

• Relevant experience and environments

• Behavioural and achievement signals

• Contextual skills

This process enables a more complete understanding of the role beyond keyword matching.

All AI processing takes place within hackajob’s controlled infrastructure.

Step 3 - Candidate Discovery

Archer searches exclusively within the hackajob community to identify professionals whose experience aligns with the calibrated role requirements.

It does not access your internal talent pools, employee records or historical hiring data.

Step 4 - Qualification

Archer evaluates whether a candidate meets the defined requirements of the specific role, based on structured profile data and evidence of experience and achievement.

This qualification takes place before any application is submitted, and entirely within hackajob’s environment.

Step 5 - Candidate Application

If a candidate chooses to proceed, they apply for the role through your standard recruitment channel.

From that point onwards:

• Your recruiters assess the application

• Your systems manage the process

• Your hiring managers make all progression decisions

Archer does not assess, rank, filter or reject applicants within your ATS. All progression decisions remain under the control of your recruitment and hiring teams.

What AI Is Used For

AI is used solely to interpret and calibrate role requirements derived from the externally exposed role feed.

This supports more accurate identification of relevant candidates and improves the quality of applications received.

AI is not deployed within your internal systems and does not interact with applicant data stored in your ATS.

Our broader AI governance principles are outlined in the hackajob AI explainability guide.

What AI Is Not Used For

Archer does not:

• Access candidate records inside your ATS

• Evaluate applicants within your recruitment systems

• Rank or score applicants on your behalf

• Make or influence hiring decisions

• Reject candidates on your behalf

Fairness and Bias Mitigation

hackajob is committed to ensuring that AI-assisted matching is fair, consistent, and free from discriminatory outcomes. The following safeguards are built into Archer’s design:

Objective matching criteria

Candidates are matched to roles based solely on skills, experience, and role-relevant criteria. Factors such as gender, ethnicity, age or educational institution are not used as matching signals. No personally identifiable information (PII), including protected characteristics, is exposed to AI models.

Bias audit and monitoring

hackajob conducts regular reviews of AI outputs to detect and reduce unintended bias. Models are version-controlled, monitored for performance drift, and evaluated continuously for fairness across candidate groups. Periodic human validation is carried out to identify and address any bias patterns.

Human decision-making retained

AI provides candidate identification and relevance signals only. All shortlisting, progression and hiring decisions remain with your team. Archer does not influence those decisions.

Compatibility with equal opportunities obligations

Archer’s architecture is designed to be compatible with employer obligations under the Uk Equality Act 2010 and equivalent legislation. Because AI is used only at the pre-application discovery stage, with no involvement in screening, shortlisting or rejecting decisions, it does not interfere with your existing equal opportunities processes or governance.

Regulatory Positioning

EU AI Act

The EU AI Act categorises AI systems used in recruitment and employment as “high-risk”. However, not every AI tool used in a recruitment context is automatically high-risk - the classification depends on the specific function performed.

Archer’s architecture is designed to mitigate regulatory exposure:

• AI is used only at the pre-application discovery stage; candidates are not assessed, ranked, or rejected by Archer

• All hiring decisions remain withyour teams

• Archer does not operate within your ATS or integrate with your internal systems

• Candidates choose independently whether to apply - no binding outcome is produced by AI

hackajob’s position: Given that Archer is limited to candidate discovery and does not produce binding outcomes affecting individuals, hackajob believes Archer operates at the lower end of the risk spectrum within the EU AI Act framework. hackajob is actively monitoring regulatory developments and will provide updated guidance as the Act’s implementing rules are finalised. 

GDPR - Automated Decision-Making (Article 22)

GDPR Article 222 gives individuals the right not to be subject to solely automated decisions that produce significant legal or similarly significant effects.

Archer does not constitute automated decision-making within the meaning of Article 22 for the following reasons:

•  Candidates are not rejected by Archer - they are shown relevant roles and choose whether to apply

• No binding outcome is produced for any individual

• Archer’s outputs are opportunities presented to candidates, not decisions made about them

• All downstream decisions (screening, shortlisting, and offer) are made by your recruiters and hiring managers.

UK Equality Act 2010

Archer does not use protected characteristics (such as age, gender, ethnicity, disability or religion) as matching criteria. Candidates are assessed solely on skills, experience and role-relevant factors. This is consistent with obligations under the UK Equality Act 2010.

United States - Regulatory Landscape

The US does not have a single federal AI law. Instead, the landscape combines long-standing federal civil rights principles with a growing set of state and local requirements. Across all of them, the core concern is the same: AI tools that make or substantially assist decisions about which candidates are screened, shortlisted or rejected must not produce discriminatory outcomes and must operate with appropriate human oversight.

Archer’s architecture addresses both concerns by design. Matching is solely based on skills and experience. And because Archer operates at the pre-application discovery stage rather than within a hiring workflow, it falls outside the primary focus of emerging AI hiring regulations, which target tools that influence decisions about applicants already in a process.

Security and Governance

Archer operates within hackajob’s enterprise security and compliance framework. The key certifications, commitments and data governance measures are summarised below.

• ISO 27001 - Certified Information Security Management System covering risk assessment, access controls, encryption, secure software development and third-party audits

• SOC 2 Type 1 - Point-in-time certification of security, availability, processing integrity, confidentiality and privacy controls.

• GDPR Compliance - Compliant with UK GDPR and the Data Protection Act 2018, and EU GDPR

• Incident Response - In the event of a personal data breach, hackajob will notify affected customers without undue delay and, where feasible, within 72 hours, in line with UK and EU GDPR requirements.

Full details of security controls and data protection measures are available at our Trust Centre: trust.hackajob.com 

Candidate Data and Consent

Candidates on the hackajob platform are informed that AI is used to match them with relevant roles. Job seekers choose how they engage with matched opportunities. Diversity data (where collected) is gathered with explicit candidate consent and is not shared with employers without the candidate’s knowledge.

Only the minimum data required to generate match explanations and role relevance is processed by AI components. Customer and candidate data are logically isolated and cannot be cross-referenced between clients.

Business Outcomes

Archer is designed to deliver measurable outcomes without requiring changes to your internal recruitment processes or governance:

• Identification of relevant, passive candidates who would not typically apply through traditional channels

• Higher interview conversion rates as a result of pre-application qualification

• Reduced time spent on low-relevance applications

• AI adoption in recruitment without the compliance overhead of ATS integration or automated screening

These outcomes are achieved without changing internal hiring processes or delegating decision-making authority to automated systems.

Conclusion

Archer provides an externally operated AI capability focused exclusively on improving candidate discovery and relevance at the top of the recruitment funnel.

It does not integrate into internal systems, does not access internal applicant data, and does not participate in hiring decisions.

This architecture enables organisations to adopt AI within recruitment while retaining full control over governance, compliance and equal opportunities obligations.