Senior data scientist, 6 years in Python and SQL, forecasting models in production.
Developed forecasting models in Python and SQL pipelines for a product team.
Model monitoring: accuracy and data drift.
BRAIN HR reads the whole CV batch against the mandate, links every shortlisted profile to its evidence and prepares the client dossier. You keep the decision.
Animated illustration with fictional data: Client request, Defensible shortlist, Client dossier.
Build your response around what matters to the client: relevant profiles, clear evidence and a dossier they can use.
The first agency to present solid profiles keeps control. Reading hundreds of CVs by hand costs days.
An unexplained score convinces neither the client nor the candidate. You need to show where the criterion is proven.
An uneven or approximate dossier weakens the relationship, even with the right candidate.
Experience, evidence and open questions, together in the dossier.
Developed forecasting models in Python and SQL pipelines for a product team.
What was your role in deploying this model to production?
Follow a client request through BRAIN HR. At every step, a concrete result your team can check.
Must-haves, preferences and exclusions are set with you before screening.
ATS, talent pool and inbound applications: duplicates merged, suspicious files isolated.
Every profile is read against the same grid; gaps become interview questions.
Developed forecasting models in Python and SQL pipelines for a product team.
What was your role in deploying this model to production?
Evidence, reservations and interview notes in your client’s format, after your review.
Ready for your review
In the product
Comparison matrix
Candidates are read side by side against the same must-haves. A missing proof is shown as such, not smoothed into a score.
Located evidence
Each selected profile keeps the exact CV sentence behind the assessment. Your client can check it; your consultant can defend it.
Experience
Kora Systems · Data scientist · 2020 — 2024
Developed forecasting models in Python and SQL pipelines for a product team.
E1 → Python · SQLWeekly reporting for the sales team, dashboards maintained in production.
Worked with the platform team to deploy forecasting models and monitor their performance.
E2 → Deployment · to confirmTo explore: personal ownership of model deployment.
Client templates
Logo, colours and layout: every dossier follows the template of the client company it is sent to. Nothing leaves before a consultant approves it.
Animated illustration with fictional data: Norvia Analytics, Kea Systems, Helix Assurance.
Demo scenario
Fictional scenario · Senior data scientist · Norvia Analytics
No. The platform reads, compares and prepares. Deciding to present a profile stays with you, after reviewing the sources.
Yes, from a template active in your workspace. Generation is validated with your templates during onboarding.
From your own sources: ATS, talent pool and inbound applications, brought together in one batch. Integrations depend on your configuration.
The criterion is marked “not evidenced” or “to confirm” and becomes an interview question. Nothing is inferred.
Yes. Periods not covered by the stated dates and overlapping experiences are highlighted as points to clarify in the interview. They do not automatically disqualify a candidate.
Within a mandate, profiles are assessed against the same agreed criteria, with must-haves kept separate from preferences. Each new mandate has its own grid: a result should always be read in the context of the client request.
Your consultant reviews the evidence, the points to confirm and the client format before deciding to share a dossier. Preparing a document does not replace that approval.
No. A typical mandate, your usual volumes and your main constraint are enough for the 30-minute demonstration. Example profiles can be used to explore the workflow.
Describe a requirement in a few minutes: volume, seniority, main constraint. No CV or candidate data is requested.