AI Solutions Engineer

I've been the analyst stuck in repetitive work. Now I build the automations that fix it.

If your team is spending hours gathering data, writing reports, or screening things manually — most of that can probably be automated. I've done it in biotech, real estate, and edtech. Working tools in days, not months.

Start a conversation See my work

Where I add value

I build automations grounded in how the work actually gets done, because I've been the one doing it. Working systems in days, not months. Then iteration based on what people actually use.

At a glance

Built tools first for myself, then my team, then as my job. Always starting with the manual work.
One automation replaced £24k/year in market research spend. Still running, two years later.

I got into this because I was the person stuck doing the tedious work. Not because it was hard, but because it was mindless. So I started building my way out of it.

Experience

Background

How I got here

Business analyst turned AI builder. I learned the hard way that you have to do the job manually first, or you build something nobody wants to use.

2022

PrecisionLife

Business analyst in biotech

Began in business analyst roles doing competitive research, partner assessments, and strategy work. The kind that went out of date the moment it was finished.

PrecisionLife

Business analyst in biotech

Began in business analyst roles doing competitive research, partner assessments, and strategy work. The kind that went out of date the moment it was finished.

PrecisionLife

Building better ways

Saw smart teams stuck in tedious tasks, so I started building automated tools. The early ones weren't great. I had to learn that without understanding how things are done manually, you build something that doesn't quite meet the need, or people find it tedious and don't use it. That lesson shaped everything after.

2023

PrecisionLife

Building better ways

2024

PrecisionLife

Automated away

The competitive intelligence automation replaced external spend and ran quarterly. Turns out understanding the problem matters more than knowing the tooling.

PrecisionLife

Automated away

The competitive intelligence automation replaced external spend and ran quarterly. Turns out understanding the problem matters more than knowing the tooling.

Fifth Dimension

Same pattern, new domains

Applied the approach at Fifth Dimension (real estate) and Multiverse (edtech): find the bottleneck, define good-enough, ship quickly, iterate with usage.

2024-2025

Fifth Dimension

Same pattern, new domains

Applied the approach at Fifth Dimension (real estate) and Multiverse (edtech): find the bottleneck, define good-enough, ship quickly, iterate with usage.

2025

Multiverse

Full-time AI specialist

Now building AI solutions full-time. Same principle: understand the manual work first, then automate what matters. The tools are better but the approach hasn't changed.

Multiverse

Full-time AI specialist

Now building AI solutions full-time. Same principle: understand the manual work first, then automate what matters. The tools are better but the approach hasn't changed.

What I've built

Real tools, real problems

Internal tools, client solutions, and sales prototypes. All built to solve a specific problem someone actually had.

Operations automation

At Multiverse, I've built and delivered multiple projects to improve workflows for coaches. The goal is to free up their time so they can focus on apprenticeships rather than admin.

Outcome

Keeps revenue per employee increasing by removing manual repetitive work

Zapier

Cassidy

OpenAI

Anthropic

Client solutions and enterprise PoCs

At Fifth Dimension, I built AI solutions for clients and proof of concept prototypes for prospective deals. That meant scoping in discovery calls, building demos, presenting them live, and answering technical questions.

Outcome

$120k+

PoCs generated $120k+ in pipeline by progressing deals to contract negotiation.

OpenAI

Anthropic

Google Gemini

Claude Code

Internal tools

The partner assessment tool at PrecisionLife automated data gathering and first line of screening for the commercial team. Structured outputs replaced hours of prep work.

Outcome

30-50%

30-50% efficiency improvement; the team spends time on decisions, not data pulls and first line screening.

Make

OpenAI

Google Gemini

Perplexity API

Results

What actually worked

I track these so I know what actually worked. Not vanity metrics.

PrecisionLife

Market research cycle

Before

Weeks of manual work

After

Automated quarterly reports

PrecisionLife

Partner assessment time

Before

Hours per company

After

Minutes per company

Fifth Dimension

PoC delivery

Before

Slide decks and promises

After

Working demos in days

£0k/year

Annual spend avoided

Market research automation at PrecisionLife

$0k+

Pipeline from PoC solutions

Fifth Dimension

0-0%

Efficiency improvement

Partner assessments at PrecisionLife

Builds

Real things I've built, explained

Walkthroughs of real projects — the process, the decisions, and what I learned along the way.

Planning Beats Syntax: How I Built This Site

How I built a production portfolio site with Next.js, TypeScript, and Tailwind CSS using AI coding tools. No formal software engineering background.

Next.jsTypeScriptTailwind CSSClaude Code

See all builds

Projects

Three projects I can walk you through

Three projects I can actually walk you through. Cost reduction, internal tooling, and a customer feature.

Cost reduction

Competitive Intelligence Automation

R&D leadership needed coverage across 8 diagnostic market segments and manual reports took weeks. Purchasing off the shelf reports would have cost £24k annually. I built a Make automation that runs quarterly, pulls Exa Search data on market segments, extracts data using Google Gemini, saves it to Airtable, and uses OpenAI models to create a executive summary report.

Outcome

£24k

£24k annual spend avoided. Delivered new business capability to generate competitive intelligence reports on demand.

Make

Exa Search API

Google Gemini

OpenAI

1Exa Search API

2Raw market data

3Gemini extraction

4Structured insights

5Airtable storage

6OpenAI summary

7Executive report

Internal tooling

Partner Assessment Tool

The commercial team spent hours gathering, structuring, evaluating prospect information. I built an AI-powered automation that takes a list of companies and prompts Perplexity to gather research information on each company, evaluates it against partnering criteria using OpenAI models, and stores everything in a structured format in Airtable for instant use by the commercial team.

Outcome

30-50%

30-50% efficiency improvement - time to carry out partner assessments reduced from hours to minutes.

Make

OpenAI API

Perplexity

Airtable

1Company list input

2Perplexity research

3Criteria evaluation

4OpenAI scoring

5Airtable output

Customer product features

Multi-Document Batch Analysis

Enterprise customers needed the ability to quickly find answers across many real estate documents at once. I shipped a customer-validated feature that parallelised question analysis, removed hours of manual review, and paired it with an explainer Loom for adoption.

Outcome

Minutes

Boosted customer satisfaction and retention by cutting customer document analysis to minutes.

Multi-LLM orchestration

Python/JSprompt libraryhandoff docs

Process

How I work

Step 1

You show me the problem

Walk me through the manual process. I need to see it, not just hear about it.

Step 2

I scope what's worth automating

Not everything should be automated. I find the highest-impact starting point.

Step 3

Working version in days

A functional prototype, not a pitch deck. You can test it with real data.

Step 4

Iterate based on usage

I watch how people actually use it, then improve what matters.

Let's talk

Start a conversation

You know your processes better than anyone. If you can walk me through the manual work, I'll build something that handles it.

Start a conversation

Reach out directly via LinkedIn or email. I respond to all messages.

Best for professional inquiries

matthew.sinnett@hotmail.co.uk

My toolkit

Automation

Make

Zapier

n8n

Cassidy

AI / LLMs

OpenAI

Anthropic

Gemini

Claude Code

Development

Python

JavaScript

Cursor IDE

Claude Code

Data

Airtable

Exa Search

Perplexity

Firebase

Common questions

Anything with a repeatable pattern: data gathering, report generation, screening, document analysis, and manual data entry. If you're doing the same steps every time, there's usually a way to automate most of it.

A working prototype usually takes days, not weeks. I ship something testable fast, then iterate based on how people actually use it. Full production-ready tools depend on complexity, but I avoid long build cycles.

Not much. Most of what I build connects to tools teams already use: spreadsheets, Airtable, Slack, email. I handle the technical setup and handoff documentation so your team can maintain it.

I pick based on the problem, not the hype. Different LLMs are better at different tasks. I've shipped production work with OpenAI, Anthropic, Google Gemini, and Perplexity, and I use whichever fits the use case and budget.

The tools will keep changing. The approach won't: understand the work, build something useful, and make sure people actually use it.