How It Works

The platform uses semantic AI, which means it understands the meaning of people’s experience rather than just scanning for exact keywords. Traditional systems often filter out strong candidates because they didn’t use the “right” words. For example, someone might have exactly the skills you need but describe them differently on their CV. A keyword-based system can miss them.

Our AI reads profiles and job descriptions more like a human recruiter would: it looks at the overall story, the context, and the underlying skills and responsibilities. That way, people with highly relevant experience aren’t filtered out just because they used different wording.

No. The AI proposes and explains matches, but every decision is made by people. The shortlist, interviews, and the final selection are always in the hands of the client, the expert, and our team. No candidate is ever rejected automatically and feedback will always be provided by the employer.

The longlist is the AI-generated set of strong candidates in response to an employer brief, each with reasoning, highlights, and open questions. An AI screening then clarifies the open points and the administrative terms - availability, rate, work model - directly with the expert. Only when everything fits does a profile move on: the shortlist is the human-curated selection of 2 to 4 profiles presented to the client.

Fairness is built into how our platform is designed. Bias can enter a recruiting system through the data a model is allowed to see and through the way profiles are presented and read. We address both with a layered approach.

1. The AI never sees what it should not weigh. Our matching agents do not receive personal attributes such as age, gender, nationality, home address, or photo. These fields are stored separately from the professional profile and are never passed into matching. The agent evaluates only what is relevant to the role: summary, experience, education, languages, certifications, and preferences such as availability, rate, and work model. An agent cannot be biased by data it does not have. This is our first and strongest safeguard.

2. We do not enrich profiles from the internet. We deliberately pull no additional data from social networks, search engines, or third-party services. The only source is what the candidate provides, by uploading a CV or entering it on the platform. Candidates stay in control of their data, and no unverified signals influence a match.

3. Agents are instructed to judge on merit only. Our agents are explicitly instructed to assess candidates solely on life sciences competencies, specialised skills, verifiable experience, and the stated requirements of the role. The prestige of an institution or the location of a former employer are not scoring criteria. Every match is transparent about which requirements were met.

4. Standardised profiles reduce bias for humans, too. Whether a candidate uploads a CV or types in their details, the information is transformed into one consistent, structured format. Every profile looks the same: same sections, same order, same level of detail. Layout, design, and writing style no longer play a role, so the hiring teams who receive a match focus on substance rather than presentation.

5. Ongoing monitoring. We treat our matching engine as a high-risk AI system under the EU AI Act and log every AI evaluation for traceability. We are committed to periodic bias reviews of recommendation patterns and to reviewing our prompts whenever the underlying model changes, so the system learns from objective outcomes rather than historical industry patterns.

No. Client briefs are never posted on a public job board. A brief is only shown to experts who have been identified by the AI matching as a strong fit, so your search stays confidential.

No. Your profile is never publicly listed and your data are protected. A client only sees your profile after the AI has proposed a match, the screening is done, and you have decided that the opportunity fits you. From that introduction on, the process is human to human.

Yes. Because the matching reads the full career history instead of job titles, it surfaces deep specialists and broad generalists alike. Experience across different roles, industries, and projects counts toward the match, not just the most recent title.

The LifescienceTalent.ai platform was developed in Switzerland and is hosted in Germany.

In Switzerland, a second job is generally permitted, but employees must respect their duty of loyalty towards their main employer.

Importantly: check your employment contract and internal policies first. If notification or approval is required, inform your employer before starting the second job, ideally in writing. Provide details such as the employer, role, working hours and expected duration.

When can the employer object? A second job may be restricted if it competes with the employer, creates a conflict of interest, harms the employer’s reputation, or affects your performance in your main job.

Are there limits on working hours? Yes. Working-time and rest-period rules generally apply across both jobs combined. In particular, employees normally require at least 11 consecutive hours of daily rest. Other limits apply to maximum weekly working hours, Sunday work and weekly rest.

Watch-out: rules can differ for public-sector employment or specific professions.