Konstantina Gianniou
Konstantina Gianniou
I'm Konstantina, a data analyst, environmental scientist and operations optimization specialist based in Thessaloniki, Greece. I combine research-grade statistics (R, mixed-effects models, regression) with hands-on operational experience: managing quality for a 25-agent operation at >94% CSAT and leading a Lean Six Sigma initiative that cut training gaps by 30%.
What I offer
White-label or co-delivered analytics: dashboards, KPI reporting, statistical analysis
Process and quality audits for contact centers and service businesses
Digital toolkits (QA, onboarding, coaching) that can be bundled or recommended
Content collaborations: guest posts, webinars, joint templates
Who I want to partner with
Contact center consultancies and training providers who need data or QA depth for clients
CX/support software vendors (helpdesk, QA, WFM tools) who want implementation, dashboards or templates for customers
Research consultancies and ecology/environmental firms that need R/statistics capacity
Accounting/IT/security service firms whose clients need reporting and SOP design
Creators and educators in customer service, HR and operations for cross-promotion
€
Εύρος Τιμών
6 χρόνια
Εμπειρία
Υπηρεσίες
Portfolio
3 έργα που υλοποιήθηκαν
Onboarding Improvement_Customer Service
New hires generate a spike of support questions in the weeks after training, costing coach time and slowing ramp-up. I used a DMAIC approach to test whether these questions were concentrated in a few topics and whether a short quiz could reduce them. I logged 529 baseline and 224 post-intervention questions across 20 topics. Five topics made up 58% of the baseline, so I designed a deliberately difficult quiz for them. Chi-square and Benjamini–Hochberg-corrected tests showed the overall topic mix changed significantly (p = 0.024) and uncovered one confirmed new training gap (Extranet account). Availability, the main quiz target, dropped from 53 to 10 questions but didn't survive multiple-comparison correction, so I reported it as a strong signal rather than proof.
Security Operations Center - Operational Audit
Security operations centers handle a constant stream of alarm signals, and most are false alarms. This R project reproduces the questions I asked as a SOC manager: which agents are genuinely slower, what drives false alarms, and which signals deserve escalation. The data is synthetic (1,200 signals, 10 agents, Jan–Jun 2025) so no real client or operational data is used. Statistical testing showed that seniority strongly affects response time (seniors about 4x faster), that false alarm rate is independent of shift (pointing to client-side causes), and that Critical Panic Button alarms are real over 75% of the time. A Welch t-test also confirmed a planted response-time decline for one agent (E08).
Impact of Dryocosmus kyriphilus on Castanea sativa leaf physiology and morphology
Completed during my internship at ELGO Demeter in cooperation with Aristotle University of Thessaloniki, this project examines how infestation by the Asian Chestnut Gall Wasp (Dryocosmus kuriphilus) relates to leaf physiology and morphology in Castanea sativa, across five provenances grown in a common garden. In R, I used ANOVA, Kruskal-Wallis, Tukey post-hoc tests, linear mixed-effects models, correlation analysis and PCA (the first three components explain about 73% of variation). The results showed that physiological traits (photosynthetic assimilation, leaf dry matter content) differed between groups more clearly than overall infestation damage did. The repository uses synthetic data, so it demonstrates the workflow rather than biological conclusions.
Διαβάστε αξιολογήσεις από πελάτες που έχουν συνεργαστεί με αυτόν τον επαγγελματία.
Διαβάστε αξιολογήσεις από πελάτες που έχουν συνεργαστεί με αυτόν τον επαγγελματία.
Δεν υπάρχουν αξιολογήσεις ακόμα..
Γίνε ο πρώτος που θα αξιολογήσει!