Adam Carnaffan / Engineer + founder

I build machines that run biology.

I design end-to-end automated molecular systems—from sample preparation through sequencing analysis—integrating assays, robotics, controls software, and bioinformatics.

See the work LinkedIn

Named inventor: US patent application · 20250116584
Education: EngSci · Machine Intelligence
Based: Toronto, Canada

01 The job

The whole system is the product.

Biology is already programmable. The hard part is making every layer work together.

I work across the seams: biology and machines, research and operations, physical constraints and software abstractions.

A protocol is only as strong as the instrument running it. An instrument is only as useful as its controls. Analysis only matters when the entire loop returns a clear, trustworthy result.

I stay with that loop—finding the brittle interfaces and turning a promising demonstration into something people can actually operate.

Layers I have personally built or integrated

Assays
Reagents
Sample preparation
Electronics
Robotics
Controls software
Sequencing
Bioinformatics

02 Selected work

Responsibilities, not just domains.

What I’ve actually built.

The useful part of working across disciplines is ownership: being able to follow a system from a biological sample to a physical mechanism to a software decision—and fix whichever layer is failing.

Kraken Sense / automated qPCR

Software, controls, and integration for an automated qPCR system.

Focus Software engineering → system integration

I joined on the software side and grew into system integration across electronics, robotics, laboratory workflows, controls, and product testing.

The job was to make the layers agree: translate a wet-lab protocol into machine states, coordinate the physical process, expose failures clearly, and return a result an operator could trust.

Kraken Sense automated molecular testing

Laboratory automation / sequencing

Nucleic-acid capture and library preparation, converted into robotic workflows.

Focus Assay methods → robotic workflows

I developed sample-preparation and sequencing-library methods, then automated them through robotics to the point of sequencing. That work crossed assay design, liquid handling, instrument control, failure recovery, and validation.

I also worked on reagent formulations and room-temperature stability for enzymes, RNA-based materials, and oligonucleotides—because automation is only useful when the biology survives the real operating environment.

Kraken Sense sequencing preparation

Magnus Biosciences / current

Closed-loop infrastructure for molecular discovery.

Focus Experimental design → next-decision orchestration

At Magnus, I am building systems that choose what to test, run experiments on automated laboratory infrastructure, analyze the result, and decide what should happen next.

The work connects experimental design, robotics, data capture, bioinformatics, and operations so that a discovery campaign is traceable and repeatable rather than a pile of disconnected research artifacts.

Magnus Biosciences

Bioinformatics R&D / edge systems

Sequencing analysis designed for constrained hardware.

Focus Biological signals → low-power analysis

I am exploring signal-processing and frequency-domain approaches to biological data that can move sequencing analysis away from GPU-heavy infrastructure and onto lower-power hardware.

The intended environment is the edge: places where a useful answer cannot depend on a data center, reliable connectivity, or abundant power.

03 How I work

The interfaces are where systems fail.

Own the interfaces.

01
Follow the whole path

Trace the sample, protocol, hardware state, control logic, data, and operator action as one system.
02
Build the failure cases

Test the contamination, drift, dropped step, bad sensor reading, and recovery path before the field does.
03
Make it operable

A system is not finished when an engineer can run it. It is finished when the intended operator can trust it.
04
Keep the loop closed

Logs, measurements, experimental outcomes, and costs should feed directly into the next decision.

04 Background

Engineering, academics, and a surprising amount of ice.

Built on first principles.

I studied Engineering Science at the University of Toronto, specializing in Machine Intelligence. Before engineering, I spent years in competitive ice dance—another activity where the details, interfaces, and failure modes are difficult to fake.

2018—2022 University of Toronto: BASc, Engineering Science
Machine Intelligence option
2018 Academic recognition: Governor General’s Academic Medal
Bronze · Uxbridge Secondary School
2014 Ontario Winter Games: Juvenile ice dance competitor
Yes, the Fiesta Tango is still on the public record.

05 Connect

Working on a hard system?

If it crosses biology, hardware, and software—especially molecular diagnostics, laboratory automation, closed-loop discovery, or edge bioinformatics—I’m interested.

LinkedIn GitHub