AT Armand The Analyst
PROFESSIONAL · PORTFOLIO

Real engagements, already delivered.

Six anonymized case studies drawn from delivered data-engineering, analytics and automation work. Names, sites and identifying numbers removed; methods and outcomes are real.

01
The REDCap platform beside a national TB active case-finding session in the field.
National-scale TB active case-finding — five REDCap databases unified into one reporting view.
Public health — national-scale TB active case-finding data platform

National-scale TB screening data platform: five REDCap databases unified into one reporting system

Goal — one trustworthy, near-real-time view of a half-million-person screening campaign

494,400
People-screened target
5 REDCap DBs
Unified into one system
7 regions
Coverage across Cameroon
~13 min
Incremental data refresh
public-healthtuberculosisREDCapdata-pipelineAirflowPython / Polars
Read the full case study →
02
A GeneXpert instrument beside its results software — the raw exports behind the ETL pipeline.
The GeneXpert instrument and its results software — roughly 1,700 raw files per run turned into one clean result table.
Diagnostics data engineering — molecular TB testing (GeneXpert MTB/RIF)

GeneXpert MTB/RIF results ETL and pooled-testing analytics

Goal — turn messy, multi-encoding instrument exports into one clean, analysis-ready result set

1,721 files
Parsed per scheduled run
133,556 rows
Consolidated result set
17 labs
Laboratories consolidated
77.5 / 84.3 min
Median time-to-result (pos/neg)
ETL pipelineGeneXpert / TB diagnosticsencoding detectionPython / Polarspooled testingbiostatistics
Read the full case study →
03
TB treatment medication and a treatment-monitoring consultation.
Reconnecting every TB treatment record to its original enrolment across six research cohorts.
Clinical research data engineering — record linkage for treatment outcomes

Probabilistic record linkage: reconnecting TB treatment records to their original enrolment

Goal — make sure every patient on treatment is linked back to their screening record

6 cohorts
Consolidated and matched
2-stage
Phone + fuzzy demographic match
0.80
Match-confidence threshold
Hourly
Automated Airflow refresh
record-linkagerapidfuzzREDCapAirflowPower BIPython
Read the full case study →
04
A health worker performing a point-of-care check at a community health centre.
Near point-of-care TB testing at community health centres — field dumps to clean, quality-scored records (representative photo).
Clinical data engineering — near point-of-care TB testing pipeline

Near point-of-care TB testing: an automated dump-to-REDCap pipeline with built-in data-quality scoring

Goal — reliable, continuously quality-checked field data from community testing sites

3 REDCap projects
Orchestrated from one pipeline
13 forms
Screening, specimen & lab
8 test types
Across 3 diagnostic categories
4-level
Data-quality priority taxonomy
data-qualityREDCapdeduplicationschedulingSQLitePython
Read the full case study →
05
A sheet of generated participant QR codes beside specimen handling in the laboratory.
Generated QR identities for every participant and specimen — from the field to the laboratory bench.
Field operations / research data systems — identity & specimen tracking

Million-scale QR identity and specimen-tracking system

Goal — give every participant and specimen a unique, scannable, collision-free identity

~2.04M codes
Generated cumulatively
371 sites
Tracked in the shared counter
3,742 sheets
Print-ready PDF batches
10 regions
Covered across Cameroon
PythonqrcodereportlabSharePoint / Graphfile-lockingfield data
Read the full case study →
06
The delivery team at work beside the Microsoft Planner board behind the automated digest.
Automated daily Microsoft Teams digests of operational progress, drawn from six Planner plans.
Operations automation — Microsoft 365, Planner & Teams

Microsoft Planner → Teams operational-reporting automation

Goal — give the team automatic, timely visibility of operational progress

6 plans
Tracked automatically
24 h
Rolling completion digest
Auto @mentions
Right people notified
SharePoint
Master list kept in sync
Microsoft GraphPlannerTeamsMSALSharePointautomation
Read the full case study →
LET'S TALK

Ready to get started?

Tell me about your project, question or goal. I reply within 24 hours, in English or French.

Get in touch