AI · Genomics · Precision Medicine

Antibioticresistancedecisionsinminutes,notdays.

BactoAI uses machine learning to predict antimicrobial resistance from bacterial genomes — before lab results come back.

Backed by clinicians and researchers across Africa & the UK
0.952
ROC-AUC on Meropenem
Internal validation
<5 min
From genome to result
vs. 48–72 hrs lab
6
Antibiotics predicted
Expanding panel
The Stakes

A silent pandemic is outpacing the drugs we have to fight it.

0.00M+
Deaths from antimicrobial resistance every year
48–72h
Traditional lab turnaround
10M
Projected annual AMR deaths by 2050
$100T
Cumulative global economic cost by 2050
Today
48–72h
Culture + susceptibility testing
→
With BactoAI
< 5 min
Genome-to-prediction
"Antimicrobial resistance is one of the top ten global public health threats facing humanity."
— World Health Organization
Built with support from world-class research institutions
Built by
Kenyatta University
Supported by
THRIVE
Grant
CDIE Catalyst
Program
East Africa Biodesign
Bootcamp
GEES
The Problem

Antimicrobial resistance is one of the greatest global health threats.

Traditional culture-based diagnostics delay clinical decisions, contributing to inappropriate antibiotic use, prolonged hospital stays, and increasing antimicrobial resistance.

1.27M+

Deaths directly attributable to AMR each year worldwide.

48–72 hrs

Traditional antimicrobial susceptibility testing turnaround.

Millions

Patients receive empirical antibiotics before lab confirmation.

Our Solution

Meet BactoAI

BactoAI analyzes bacterial genomic data using machine learning to predict antibiotic resistance before conventional laboratory testing is completed.

STEP 1
Patient

Clinical presentation & bacterial infection identified.

STEP 2
Sample

Bacterial isolate collected from the patient.

STEP 3
Sequencing

Whole genome sequence generated (FASTA / FASTQ).

STEP 4
BactoAI

ML models analyze genomic features.

STEP 5
Prediction

Resistance profile & confidence per antibiotic.

STEP 6
Treatment

Clinician makes an informed prescribing decision.

Live Product Demo

Try it yourself. Watch a genome become a treatment recommendation.

Pick a sample isolate, then click Analyze sample to simulate the BactoAI workflow — upload, pre-processing, model inference, and clinician-ready report.

  • Drag-and-drop FASTA / FASTQ upload
  • Per-antibiotic resistance & confidence
  • PDF clinical reports
  • Audit-ready result history
1

Select a sample to analyze

Four de-identified bacterial whole-genome samples from our validation set.

2

Run the prediction

app.bactoai.com/predict
Patient ID
PT-00184-KE
Ready
isolate_A12.fasta
3.2 MB · WGS · K. pneumoniae · Ready to analyze
Results will appear here after analysis. Change the selected sample above to see different genomic resistance profiles.
3

Request a demo on your own isolates

Send us a request and our team will run BactoAI against your genomes with you.

Validation

Evidence-based development.

Our machine learning models are developed on curated bacterial genomes and rigorously evaluated. Prospective clinical validation is ongoing with partner laboratories.

  • Trained on curated public bacterial genome datasets
  • Internal cross-validated performance metrics
  • Clinical validation studies in progress
1,800+
Training bacterial genomes
3
Antibiotics currently modeled
0.952
ROC-AUC — Meropenem (internal)

Reported figures reflect internal model development. Performance on prospective clinical isolates will be reported upon completion of ongoing validation.

Ready to transform AMR diagnostics?

Bring genome-driven decisions into your hospital or lab.

See a live demo of BactoAI on your own isolates, or talk to us about launching a pilot in your region.