Our Product Portfolio

Products in the Pipeline

High-conviction products targeting global markets — biosignal intelligence, ML reliability tooling, and LLM evaluation. Built on real research.

All products are currently under development. Join waitlists for early access.
Global Products

High-Conviction Products

Targeting hard-currency B2B and developer customers, deliverable as API/SaaS globally.

In Development — Join the waitlist to get early access
GlobalIn Development

BioSignal API

"Stripe for physiological signals"

A REST API that takes raw ECG/EDA/PPG waveforms and returns cleaned signals, HRV features, arrhythmia flags, and calibrated uncertainty estimates — all in one call.

View details

Key Features

  • R-peak detection & HRV (SDNN, RMSSD, pNN50)
  • EDA phasic/tonic decomposition
  • Arrhythmia screening with calibrated confidence
  • PPG-derived SpO2 estimation
  • Batch processing with webhook callbacks
  • FHIR-compatible output format

Tech Stack

PyTorchNeuroKit2FastAPIAWS LambdaPhysioNet

Roadmap

  • 1Q1 2026: Private beta
  • 2Q2 2026: Public launch
  • 3Q3 2026: Regulatory guidance docs
Planning — Join the waitlist to get early access
GlobalPlanning

CalibKit

ML Calibration & Uncertainty Toolkit

An open-source Python library + paid hosted dashboard for model calibration, uncertainty quantification, abstention gating, and production drift monitoring.

View details

Key Features

  • Temperature scaling, Platt, & isotonic calibration
  • Expected Calibration Error (ECE) tracking
  • Confidence-gated abstention framework
  • Covariate shift & concept drift detection
  • Reliability diagram visualizations
  • CI/CD integration via GitHub Actions

Tech Stack

PythonPyTorchscikit-learnReactFastAPI

Roadmap

  • 1Q2 2026: OSS library release
  • 2Q3 2026: Hosted dashboard beta
  • 3Q4 2026: Enterprise tier
Idea Stage — Join the waitlist to get early access
GlobalIdea Stage

EvalForge

LLM Evaluation & Calibration Platform

A hosted eval harness for LLM-based systems — automatically measuring calibration, factuality, and abstention rates on your domain-specific test sets.

View details

Key Features

  • Automated ECE computation for LLM outputs
  • Verbalized vs implicit confidence comparison
  • Abstention rate tracking across prompt variants
  • Dataset versioning & reproducibility
  • Slack/email alerting on calibration regression

Tech Stack

PythonAnthropic APIOpenAI APIPostgreSQLReact

Roadmap

  • 1Q3 2026: Prototype
  • 2Q4 2026: Design partners
  • 3Q1 2027: Launch

Want to Learn AI from the People Building It?

We run focused training programmes in biosignal AI, ML calibration, and LLM evaluation for engineers, researchers, and students.

Partner on a Product

Are you a digital health startup, wearable maker, or AI company? We're looking for design partners and pilot customers across all products above.

Become a Design Partner