ResistAI Assistant
AI-powered Q&A
🤖
Hello! I'm your ResistAI assistant. Ask me anything about antibiotic resistance prediction, how to use the app, or interpreting results.
Just now
Biomedical AI · Clinical Decision Support · AMR Combat

Stop guessing.
Predict resistance
in seconds.

AMR kills 700,000 people yearly. Doctors wait 48–72 hours for lab results. ResistAI gives clinicians resistance probabilities for 12 antibiotics — trained on 10,000+ clinical cases — in under 2 seconds.

⚠️700K+Deaths/yr from AMR
⏱️48–72hLab wait time
⚡<2sResistAI prediction
🎯80%+Model AUC accuracy
◆

Species-aware prediction

Resistance probabilities tailored to each organism's known clinical patterns across 12 antibiotics.

◈

Explainable AI (SHAP)

Plain-English explanation for every result — not just scores, but the exact factors driving each prediction.

📄

PDF clinical report

One-click export of a complete clinical summary with patient data, resistance chart, and recommendations.

📈

Patient timeline

Track resistance evolution across multiple visits. See if a patient is developing resistance to key antibiotics.

⚖️

Compare isolates

Run two predictions side-by-side to see exactly which antibiotics changed between samples.

↩️

Lab result feedback

Submit actual lab outcomes after predictions to track model accuracy and build a learning feedback loop.

resistai — example prediction output
$ predict --species "E. coli" --ward "ICU" --prior "Ampicillin" --stay "long"
Analyzing resistance patterns... done in 0.8s · Confidence: 87%
Meropenem
8%[EFFECTIVE] ₹ Low cost
Gentamicin
42%[UNCERTAIN] ₹ Low cost
Ciprofloxacin
65%[RESISTANT] ₹₹ Med
Ampicillin
88%[RESISTANT] ₹ Low cost
✓ AI: Prior Ampicillin exposure is the strongest resistance driver (31% SHAP influence).
🧬
Helps visualize how bacterial resistance connects to different antibiotics, enabling better treatment decisions.

Resistance predictor

Enter patient and isolate data to get antibiotic resistance predictions with AI explanations and cost context

01Bacterial isolate
02Clinical risk factors
03Prior antibiotic exposure
Select all recently administered antibiotics
◉
Results will appear here
Fill in the form and click "Predict resistance" to get resistance probabilities with AI explanations and cost context.
Example output:
● Meropenem → 8% resistance · ₹ Low cost
● Gentamicin → 42% uncertain · ₹ Low cost
● Ampicillin → 88% resistant · ₹ Low cost
🧬 Resistance Gene Network Gene ↔ Antibiotic relationships
Helps visualize resistance pathways between bacteria and antibiotics.
🔍 Visual explanation of prediction

Compare two isolates

Run two predictions side-by-side to see which antibiotics changed between samples

Sample A
Prior antibiotics (Sample A)
Sample B
Prior antibiotics (Sample B)

Patient resistance timeline

Track how resistance evolves across multiple visits for a specific patient

📈
No patient selected
Enter a Patient ID and click Load. Run predictions with a Patient ID from the Predict tab to start building a timeline automatically.

Resistance dashboard

Aggregated patterns · Ward heatmap · Trend analysis

Updated: just now
1,182Total predictionsAll time
38%High resistance cases↑ vs last month
83.4%Model accuracyAUC on test set
12Lab confirmationsFeedback loop
Ward resistance heatmap% resistant per antibiotic per ward
Resistance rates by antibiotic% resistant
Resistant ≥65% Uncertain Effective
Top organismsBy volume
Resistance trends · 7-dayTop 3 species
Recent predictionsLatest 5
IDSpeciesWardBest optionResistanceTime

Prediction history

Complete log · Lab feedback · Export

0 predictions
All predictions—
IDPatientSpeciesWardBest antibioticResistanceLab resultTime
No predictions yet