Private AI for HR and recruiting
A dedicated model can summarize resumes and answer HR policy questions without a third-party AI API seeing candidate and employee records. Resume screening is one of HR's most common AI uses and one of the most sensitive: a stack of resumes holds names, addresses, employment history, sometimes photos, and signals about age, nationality, disability or family status that a model can infer even when the candidate never disclosed them.
Not legal advice. Automated decision-making in hiring is a live regulatory area in multiple jurisdictions, and the specific rules depend on where you operate and where your candidates are. This post describes an infrastructure choice, not a compliance opinion. Check with your own employment counsel before using AI in any hiring decision.
Why HR data is a different category
Most business documents are sensitive because they're confidential: a contract, a financial forecast. HR data is sensitive for a different reason: it's about identifiable people who didn't choose to be your customer or counterparty. In many jurisdictions, employment data sits closer to the categories that get extra regulatory attention. A resume alone can reveal a candidate's age (graduation years), possible disability (an employment gap explained honestly), or national origin, none of which the candidate necessarily intended to disclose as a factor in the decision. Feeding that into a general-purpose AI product with unclear retention and no HR-specific handling puts those inferences in a vendor's systems.
Employee data compounds the exposure further. Performance reviews, disciplinary records, and compensation history are the kind of information HR treats as need-to-know internally, restricted to the employee's manager and HR itself. Running an AI tool over that data through a shared third-party endpoint puts it through infrastructure with a much wider blast radius than the access controls your HR system already enforces.
What running the model yourself changes
The setup looks like any self-hosted assistant: an instruct model wired to the workflow that needs it, such as summarizing a resume against a job description, structuring interview notes, or answering a manager's question against review history. On a dedicated DGX Spark, resumes, reviews and disciplinary notes go to your own model on a machine assigned only to your organization, not to an AI vendor's API. There's no question of which plan tier your recruiter is logged into or whether it trains on submitted content.
You still need a DPA, with the hosting provider instead of an AI vendor: the Spark is a GPUwerk-run machine (its own in Czechia, or a partner's in the country the console shows), and GPUwerk acts as your processor under its published DPA.
That doesn't resolve the harder question, which is whether AI should be part of the hiring decision at all, and if so, how. Screening resumes for keyword match against a job description differs from ranking or rejecting candidates automatically, and the second is where automated decision-making rules in hiring start to apply in various jurisdictions. Keep a human reviewing and making the actual decision; use the model to summarize and surface, not to decide.
A reasonable starting scope
Start with the lower-stakes half of the workload: summarizing resumes into a consistent format for a recruiter to read faster, or answering an HR generalist's question against policy documents (this overlaps with a general internal knowledge base, if HR policy is already indexed there). Save anything that touches a ranking or pass/fail judgment on a specific candidate for later, once you've worked out with counsel what role AI is allowed to play in that decision under the rules that apply to you.
Wherever the model touches employee-identifiable data, apply the same access boundary the underlying HR system already has. A self-hosted model doesn't inherit your HR platform's permission model. Someone has to wire that access control into the front end, or the model becomes a shortcut around restrictions that took real effort to set up.
Retention is part of the design, not an afterthought
A cloud AI product's default retention window is built for the vendor's general use case, not for HR's specific obligations around candidate and employee records, which often carry their own retention and deletion rules independent of anything AI-related. Running the model yourself doesn't solve retention either: someone still decides how long the HR tool's transcripts, summaries and chat logs are kept, and deletes them on schedule. The advantage is that you set that schedule on your own instance, with no AI vendor default to override and no support ticket to file when a candidate asks what happened to their data.
Rejected-candidate data is worth deciding on explicitly. Some organizations keep resumes and screening notes for a defined period in case a role reopens; others delete on rejection. Apply your existing HR system's schedule to everything the AI tool generated from that data (summaries, extracted skills lists, screening notes), so those artifacts don't outlive the source record.
One Spark covers most HR teams
An instruct model in the 20B-70B parameter range handles resume summarization and internal HR Q&A within a single Spark's 128 GB of unified memory, alongside the retrieval layer over policy documents. HR data volumes are modest next to a company-wide knowledge base, so one node at $0.79/hour in EU-Central is usually enough.