Skip to content
MBiotech.
Get in touch

Biotechnology · Life sciences · Applied ML

Mat Beale.Science, software and sales.

I work across biotechnology, life sciences and health tech, building products and prototypes and taking on selected consulting and contract projects. I recently achieved my master's in biotechnology with a distinction. This site is the work.

Get in touchSee the work

Available for full-time roles and selected freelance projects.

Mat Beale working at the overlap of science, product and sales.
Presales & demosGo-to-marketWorking prototypesApplied ML

Services

Services I can provide

Every engagement is scoped, priced and tied to deliverables before it starts.

01

Presales & solutions consulting

Demos, POCs and hands-on help with sales and account management, without a full-time hire.

Contract or freelance · Per deal or fixed scope · day rate

02

Commercialisation & go-to-market

Positioning, pricing and a first sales process for products that are technically strong and commercially early.

Strategy + execution · Typically 2 to 6 weeks · fixed fee

03

Applied ML & prototype development

Proof-of-concept models and full-stack builds that show what your data can do, with explainability built in.

Build · Typically 4 to 8 weeks · fixed fee or day rate

Projects

Selected work

A sample of shipped products, applied research and consulting-grade analysis.

Research workspace · In previewOngoing

Drybench

A connected dry-lab workspace for bioscience: nine research tools for proteins, molecules, trial pipelines, patents, bench calculations, sequences, assays, plasmids and bioreactors, joined by a Bench that keeps each study together.

  • Building a research product end to end, from the tools to the workspace that holds the work
  • Methods stated on the page: the algorithms, assumptions and limits behind each answer
  • Keeping the context behind a result: inputs, method versions and qualifications
Read the work
Live productOngoing

FitTools

A live, public health and fitness platform: 43 calculators, a gym tracker and a food diary, with every formula cited to its published source.

  • Building and operating a product end to end
  • SEO architecture
  • Health content cited to peer-reviewed research
  • Product decisions held to an evidence rule
Read the work
Applied ML research2025 to 2026

AI/ML Diabetic foot ulcer classification

Does a classifier trained on photographs from one source still work on photographs from another? It does not, and the direction of failure flips depending on which way you cross.

  • Cross-source evaluation design
  • Data provenance auditing
  • Calibration and operating point analysis
  • Negative controls and rejected alternatives
  • Reproducible experiment scripting in PyTorch
Read the work
Clinical AI concept2026

WaitWise

An AI clinical decision support concept for NHS elective surgery prioritisation, designed with Rajdeep Kundu during the MSc: multimodal machine learning with explainability, governance and a phased route into the NHS.

  • ML system design: BioBERT note embeddings fused with XGBoost, explained with SHAP
  • NHS governance literacy: NICE ESF, UK MDR, equity audits, drift monitoring
  • Choosing methods on explainability and governance, not just accuracy
  • Presenting technical work to a mixed clinical and commercial audience
Read the work
Venture pitch2025

InfectiPatch

An investor-style pitch and scientific poster for a microneedle biosensor patch that monitors post-surgical infection through interstitial-fluid biomarkers.

  • Framing a biotech concept as an investable proposition
  • Market sizing, revenue modelling and route to market
  • Regulatory and evidence planning for a connected medical device
  • Scientific communication for investors and for a technical audience
Read the work
All work

Apps

Drybench: dry-lab tools for biology and drug development

Apps for proteins, molecules and clinical trials, built on public scientific data, and for the bench: stocks and dilutions, sequences, dose-response curves, plasmid maps and bioreactor runs. Each names the source and licence beside its numbers.

See Drybench

Why me

Background across science, software and sales

Science

A master's in biotechnology from the University of Reading, with a distinction, and an applied machine-learning research project in medical imaging, with clinical advisory input from an NHS orthotics consultant. I can read the literature, follow the methods and work directly with scientific teams.

Software

Full stack developer before I moved into presales, and co-founder of a B2B infrastructure platform from 2017 to 2021, run part-time alongside the day job. I build working products and run them in production.

Sales

Four years in a commission-based presales role at an enterprise data discovery company, 2018 to 2021: over 100 demonstrations and more than 20 proof of concept evaluations, on deals from £30k to £300k.

Next step

Get in touch.

Want to talk about a role, a project or any of the work here? Get in touch and I will reply within a working day.

Get in touchm.beale@icloud.com

Available for full-time roles and selected freelance projects.