Selected case studies

Problem

Candidate evaluation was inconsistent and manual resume screening was eating up time.

Solution

Built a Next.js app with four AI modules: resume evaluation, interview planning, transcript analysis, and structured reporting.

Impact

15–20 minutes saved per candidate and a unified evaluation standard across the team.

Problem

Paid-social leads for high-volume tutor hiring dropped into a Google Sheet, and everything after that was hands — messaging each person, chasing a CV, deciding alone who was worth an interview.

Solution

An n8n bot that picks up new leads, knocks out anyone failing the hard criteria, asks the rest for a CV on WhatsApp, scores it with an LLM, and books the ones who pass onto a group interview.

Impact

Lead → HR interview conversion went from 16.7% to 26.0% (+9.3 pp).

Problem

Almost half the new joiners in Mexico, Colombia and Argentina left early, assessment was subjective, and the programme fell apart the moment hiring scaled.

Solution

Rebuilt the programme day by day and built the infrastructure underneath it: exams and a scoring model that measures how someone works, a programme database, and an LMS written from scratch.

Impact

Probation → live sales floor conversion went from 19% to 34% (+15 pp).

Problem

Decisions were made on instinct — metrics lived in scattered exports, and nobody separated the resignations we regretted from the ones we did not.

Solution

Built people analytics from nothing: dashboards for attrition, regrettable versus non-regrettable exits, funnel conversion, LTV and payback, fed over APIs and deployed on Vercel.

Impact

One source of truth the CEO, HRD and team leads use every week.

Problem

Gift coordination was fully manual — slow, opaque, and easy to forget.

Solution

Built a Telegram bot that triggers 10 days before an employee's birthday, lets them pick from curated options, and gives HR a live status dashboard.

Impact

Zero manual coordination, works across time zones.

Problem

Action items discussed in Telegram rarely made it into the project tracker.

Solution

Make.com automation: text or voice message → AI pulls out the task → creates it in Asana with the right assignee and project.

Impact

Tasks land in the tracker seconds after the message is sent.

Problem

Without regular sentiment data, HR decisions were always reactive.

Solution

Automated pulse survey distribution via Slack — scheduled sends, response collection, and an analytics dashboard built with Google Apps Script.

Impact

Early detection of team issues and people decisions backed by data.

Problem

Onboarding was a black box: information got lost between pre-boarding and month one, and relocating employees had no clear path.

Solution

Designed a 5-stage journey map (Pre-boarding → Week 1 → Month 1 → Month 3) with owners and metrics per stage, plus a separate relocation track.

Impact

Clear ownership across HR, managers, and IT.

Problem

Early attrition was coming from undetected mismatches and unclear expectations during probation.

Solution

Built a manager toolkit with stage-by-stage checklists, 1:1 templates, progress criteria, and red flag indicators for the 30, 60, and 90-day marks.

Impact

Fit issues surface earlier, check-in rhythm is consistent for every new hire.