Writing

Thoughts on Building Software

Deep dives into AI, automation, and software engineering. Lessons learned from building real projects.

LatestAI Engineering1 June 2026

AI Engineer World's Fair 2026: what actually mattered

A long, opinionated recap of the themes that defined the AI engineering field at the start of June 2026. MCP as the universal skill protocol. Multi-provider gateways as table stakes. Voice loops as a serious production stack. Evals as code becoming mainstream. The quiet rise of local-first assistants. And what I am shipping in response.

Read article
AI EngineeringJun 2026

NVIDIA at Computex 2026: what the open-source AI engineer actually needs to know

Jensen Huang opened GTC Taipei at Computex 2026 with the largest single-keynote product drop NVIDIA has done in years: Vera Rubin NVL72 in full production, the new Vera CPU, BlueField-4 STX, RTX Spark, DGX Station GB300, Cosmos 3, Nemotron 3 Ultra, Jetson Thor with Isaac GR00T, plus the Agent Toolkit and OpenShell runtime. This is the recap for an open-source AI engineer trying to decide where to put the next six months of effort.

13 min readRead
Self-hostingMay 2026

Why I prefer ZimaOS for a home NAS

There are five reasonable choices for a home NAS operating system. I run ZimaOS on my own box. Here is why it beat the others for me, and the kind of person it does not suit.

8 min readRead
F1May 2026

Hamilton vs Verstappen, by the numbers: where the rivalry sits in May 2026

Two drivers, two eras, one paddock. Lewis Hamilton's switch to Ferrari this season has put him in the same race as Max Verstappen more often, and the head-to-head numbers have caught the attention. We pulled the career figures, plotted them, read them honestly, and looked at where the next twelve months of races could shift them.

15 min readRead
AIMay 2026

Microsoft pulls back on AI tooling for staff: when the bill outgrows the engineer

Microsoft has reportedly cancelled the bulk of its direct Claude Code licences for staff and is redirecting engineers to GitHub Copilot CLI, six months after first rolling Claude Code out internally. The story sits inside a wider thread: at the unit-economics level, AI tooling is now expensive enough that even the people selling it are choosing carefully where to deploy it. The numbers, the strategy, and the implications, in detail.

13 min readRead
F1May 2026

Lewis Hamilton on the Ferrari podium: anatomy of a Montreal fightback

Hamilton's second place at the 2026 Canadian Grand Prix is more than a podium. It is the first weekend in Ferrari red where he visibly drove the car the way the car wanted to be driven. Here is what changed in the data, what he said on the radio, what the team did differently between sessions, and why the Ferrari pit thinks the next five races look different.

12 min readRead
F1May 2026

2026 Canadian Grand Prix: Antonelli wins his fourth in a row, Russell retires from the lead

Kimi Antonelli took his fourth consecutive Formula 1 victory at Montreal on Sunday, but the race the paddock will talk about for a fortnight is the one his team-mate George Russell led for thirty laps before a power unit failure ended his afternoon. Lewis Hamilton, for Ferrari, took second. Max Verstappen completed the podium. Here is the full race, lap by lap.

14 min readRead
AIMay 2026

Google I/O 2026: Gemini 3.5, Spark, Omni, and the year Google bet on agents

Google's I/O 2026 keynote landed three things at once. A new model family in Gemini 3.5. A general-purpose agent in Spark. And an Omni model that pulls images, video, audio and text under one roof. Here is what shipped, what is still trailing, why the lineup tells you where Google thinks the next twelve months go, and what to do about it if you build with the Gemini API.

18 min readRead
AIMay 2026

Google I/O 2026 preview, Gemini 4, Proactive Assistance, and 'Remy'

Google I/O 2026 lands on May 19-20, eight days away. Gemini 4 is the headline expectation, with native image and video generation and a step change in reasoning. Beyond the model, there are real signals on Proactive Assistance and an agentic Gemini called 'Remy'. Here is what is expected, what is plausible, and what to actually watch for.

8 min readRead
AIMay 2026

The state of AI, the week everything shifted (May 4-11, 2026)

Looking back at one week, May 4 through May 11, 2026, and the AI industry feels meaningfully different. GPT-5.5 Instant, Claude Opus 4.7's full rollout, Manus Wide Research, $5.5 billion in funding on a single day, expanded US government oversight, and the Google I/O countdown. Here is the through-line, and what it tells us about where the next twelve months go.

10 min readRead
AIMay 2026

Android 17 and Gemini, what actually changes for app developers

Android 17 lands at Google I/O on May 19, and the AI integration goes beyond skin deep. Gemini becomes a system service, App Actions become AI-mediated, and Proactive Assistance reframes notification ranking around Gemini context. Separately, Google is launching Aluminium OS, a desktop OS for PCs built on Android 17. Here is the developer-side read.

7 min readRead
AIMay 2026

Gemini 4 vs GPT-5.5 vs Claude Opus 4.7, state of the frontier in May 2026

Three frontier models, three vendors, three meaningful product strategies. Claude Opus 4.7 leads on hard reasoning (87.6% SWE-bench Verified). GPT-5.5 Instant leads on consumer breadth (900M weekly users). Gemini 4 (expected next week) is the integration play. Here is the honest comparison, with corrected benchmarks, current pricing, and a recommended-default for each use case.

9 min readRead
AIMay 2026

The four autonomous-agent platforms that matter in May 2026, compared head-to-head

Manus Wide Research, OpenAI Operator, Claude Agent SDK, and Google's Chrome Auto Browse (Project Mariner's successor, after Mariner was discontinued May 4) are the four serious autonomous-agent platforms in production as of May 2026. They have very different architectures and very different sweet spots. Here is the honest comparison after a month of running real tasks through each.

9 min readRead
AIMay 2026

Manus 1.6 Max + Wide Research, autonomous agents finally go wide

Manus shipped two major upgrades in the last fortnight, the 1.6 Max architecture for deeper research, and Wide Research, where multiple Manus agents collaborate as sub-agents on the same task. Combined with Web App Builder and the Stripe + SEO integrations, this is the most consequential agent release of May 2026.

8 min readRead
RecommendationsMay 2026

A good friend in logistics, Uneek Group

I have been working alongside Uneek Group for the last two years through the company I'm on contract with, and they have been a long-standing partner there for over thirty years. This is a personal write-up of why that partnership has lasted, what they do, and how to get in touch if you need a logistics team that actually feels like a friend on the inside.

7 min readRead
F1May 2026

Kimi Antonelli: rookie last year, championship leader this year, what to expect next

P2 at Australia, then three consecutive wins at China, Japan, and Miami. Seven podiums in his second F1 season, and the youngest championship leader in the modern points era. Kimi Antonelli's start to 2026 is statistically the strongest sophomore campaign in Formula 1 history. Here is the data, the comparison, and what Imola says about whether he can hold it.

10 min readRead
AIMay 2026

Google AI Ultra Lite, where does the new tier fit?

Google is preparing an AI Ultra Lite plan to slot between $20 Pro and $250 Ultra. The expected price band is $50-80 per month. Here is who it is for, how it compares to a stacked ChatGPT Plus + Claude Pro subscription, and whether it is worth replacing what you already have.

6 min readRead
AIMay 2026

GPT-5.5 Instant, OpenAI ships a smarter, terser ChatGPT default

OpenAI rolled out GPT-5.5 Instant on May 5, replacing GPT-5.3 Instant as the default ChatGPT model. The pitch, 52.5% fewer hallucinations on high-stakes prompts, more concise responses, and full conversation memory. Here is what is actually new, what shows up in API benchmarks, and what to do if you build on the platform.

7 min readRead
RecommendationsMay 2026

A year with EV Cargo, a personal recommendation

After a year working alongside EV Cargo on the supply-chain side of my day job, I want to write something honest. This is a personal recommendation, not a paid one, just an account of why this partnership has worked, what they actually do, and how to get in touch if it sounds useful.

6 min readRead
F1May 2026

Is McLaren back? Reading the 2026 recovery in the lap-time data

McLaren defended their constructors' title from 2025 with a slow start. Then the Miami sprint flipped the narrative, Norris took the win, Piastri second, both ahead of every Mercedes. The main race told a different story. Here is what the lap-time data shows about whether the reigning champions are back, and what it means for the rest of the season.

9 min readRead
AIMay 2026

Claude Security beta, Anthropic enters the vulnerability-scanning market

Bundled with the Opus 4.7 release, Anthropic shipped Claude Security in public beta for Enterprise customers. It does code vulnerability scanning, with Claude-generated patches proposed inline. Here is how it compares to Snyk, Semgrep, and GitHub Advanced Security, and whether it is good enough to replace what you are running today.

7 min readRead
AIMay 2026

Claude Opus 4.7, what's actually new under the hood

Anthropic shipped Claude Opus 4.7 generally available across all surfaces in mid-April 2026, and the rollout finished landing this past week with Amazon Bedrock and Microsoft Foundry. The marketing pitch is 'better at hard coding tasks'. The reality is more interesting, 87.6% on SWE-bench Verified, high-resolution vision, agentic security scanning, and a 1M context window at standard pricing. Here is what changed and what I have seen in production.

8 min readRead
AIMay 2026

US government AI oversight expands, Google, Microsoft, xAI now in the testing programme

The Trump administration moved further into frontier-AI oversight last week. Google, Microsoft, and xAI agreed to provide early model access to the Commerce Department for national-security testing, joining an existing framework with OpenAI and Anthropic. Here is what is actually being tested, what the policy means for AI deployment, and what it does not change.

7 min readRead
Open SourceMay 2026

FUNDING.yml as a charity redirect: a pattern for OSS maintainers

GitHub puts a Sponsor button on every public repo by default. If you do not want to accept personal sponsorships, point the button at charities instead. Four lines of YAML, no money flows to you, and visitors get to do something kind on the way out.

5 min readRead
AIMay 2026

AI dealmaking day, OpenAI Deployment Co, Anthropic's deployment JV, and the May 4 pivot

May 4, 2026 was the most concentrated AI-dealmaking day of the year so far. OpenAI's Deployment Company raised $4 billion from outside investors at a $10B valuation. Anthropic announced a $1.5 billion enterprise-services JV with Goldman, Blackstone, Hellman & Friedman, and Apollo. Both labs are pivoting from selling models to selling deployment services. Here is what is actually happening and what it means for the rest of us.

8 min readRead
EngineeringMay 2026

Migrating Next.js 14 to 16 in a working production app

Next.js 14 had six high-severity advisories that needed clearing. The fix-in versions said 15.5.15 and above, but npm only ships up to 15.5.9. So the upgrade was 14 to 16, two majors in one push. Here is the diff that worked, the one error I hit, and what I would tell anyone facing the same jump.

9 min readRead
SecurityMay 2026

From xlsx to exceljs: when the advisory says "no fix available"

The xlsx package on npm has known high-severity vulnerabilities. The Dependabot fix says "no fix available." The reason is structural, and the right move is to switch packages. Here is the migration I shipped, what broke, and what I learned about supply-chain hygiene.

7 min readRead
F1May 2026

Lewis Hamilton's first four races in red: what the data says about Ferrari 2026

Four races into year two at Ferrari, Lewis Hamilton sits joint-fifth in the drivers' championship on 51 points, with a podium at China already in the bag. The data tells a more interesting story than the headline standings. Here is what the lap-time gaps, qualifying margins, and race-pace deltas say about where the seven-time champion actually stands.

10 min readRead
EngineeringMay 2026

Why I open-sourced 12 repos: leverage maths for a solo engineer

Eighteen months, 12 public repos, honest accounting. What each one cost in time, what each returned in leads, clients, and compounding learning. The maths behind publishing instead of hoarding.

13 min readRead
F1May 2026

F1 2026 mid-season: championship maths after the Middle East round cancellation

With the Middle East round gone, the 2026 calendar runs 23 rounds and the points maths shift. How the cost cap absorbs a missing round, and what the compressed schedule means for the title fight.

12 min readRead
AIMay 2026

Building agent-orchestrator: durable workflows, deterministic replay, and the BullMQ decision

What it actually takes to make multi-agent workflows durable: Postgres as the source of truth, deterministic replay for debugging, BullMQ for execution, and the lessons from the first 18 months.

16 min readRead
InfrastructureMay 2026

Terraform vs Pulumi vs SST in 2026: a practical comparison for solo engineers

Three infrastructure-as-code tools, each with a real niche. A comparison of language model, state, cloud coverage, blast radius, and learning curve, and when each one is the right call.

15 min readRead
AIMay 2026

SarmaLink-AI failover engine: how zero-downtime LLM rotation actually works

A deep dive into the failover engine inside SarmaLink-AI: provider weighting, latency budgets, cooldown windows after 5xx errors, and how zero-downtime rotation stays transparent to the caller.

14 min readRead
OpinionMay 2026

If Meta buys Manus, the agent market is over (for startups)

Hot take: an actual Meta-Manus tie-up would close the standalone agent-product window for startups. Distribution + compute + capital is unbeatable. Here is the case.

9 min readRead
OpinionApr 2026

If Meta acquired Manus, what would change? A speculative breakdown

Rumours that Meta is interested in Manus circulated in April 2026. Speculative, but worth considering. Here is what each side gets, what each loses, and what it would mean for the agent market.

11 min readRead
F1Apr 2026

F1 2026 Middle East round cancelled: what changes, and the tech that absorbs a calendar shift overnight

The Middle East round was pulled from the 2026 calendar at short notice. Twenty-three races, not twenty-four.

11 min readRead
OpinionApr 2026

Year on IONOS: still right for my use case in 2026

A year after consolidating to IONOS. Renewals went smoothly. Email forwarding still does what it should. The DNS UI is still ugly but works. Recommendation stands.

7 min readRead
AIApr 2026

AI agents in 2026: where the actual ground is

A grounded look at where general-purpose agents actually deliver in 2026, with benchmark numbers and a strong opinion on which products are real and which are demo-ware.

12 min readRead
CloudApr 2026

I self-hosted my side projects for a year. Here is what the spreadsheet said.

Twelve months running my side projects on Hetzner instead of Vercel. Total cost, total downtime, total time spent on operations. The numbers were not what I told myself they would be.

12 min readRead
DevelopmentApr 2026

Building a scalable SaaS platform in 2026: the boring choices that win

Five SaaS products across the last three years. The architecture choices that scaled gracefully and the ones that did not. The pattern is conservative.

14 min readRead
CloudApr 2026

Self-hosting vs Vercel in 2026: the real cost of "I will run it myself"

I migrated a side-project off Vercel onto a Hetzner box for six months, then migrated back.

13 min readRead
CloudMar 2026

Cloud migration in 2026: lift-and-shift, refactor, or rebuild

I have helped four UK SMEs migrate from on-prem or shared hosting to cloud over the last 18 months. The pattern of which strategy actually delivers the savings.

13 min readRead
AIMar 2026

Manus vs Cursor vs Devin in 2026: which AI coding agent actually ships your code?

Three months running real client tasks through Manus, Cursor Composer, and Devin. The data shows where each one wins and where each one fails. Spoiler: it is not a one-tool answer.

13 min readRead
SecurityMar 2026

Secure API architecture in 2026: the controls that actually stop incidents

I have audited 12 SaaS APIs in 2025 and 2026. The same five controls keep being missing. Get them in place and 80 percent of incident risk is gone.

13 min readRead
OpinionMar 2026

Why I stopped using ORMs in 2026 (mostly)

Three years of Drizzle, Prisma, and TypeORM. I keep coming back to hand-written SQL with a thin query layer. The data shows where the ORM tax actually hurts.

10 min readRead
CloudMar 2026

Cloudflare Workers vs Vercel Functions in 2026: a real-world cost and latency breakdown

I run the same Next.js workload on Vercel and a hand-rolled Workers stack. The pricing model is the most important difference, not the latency. Here are the numbers.

14 min readRead
F1Mar 2026

Every F1 circuit on the 2026 calendar, by the numbers

Track length, lap record, elevation, downforce profile and DRS zones for all 23 circuits on the revised 2026 calendar. The data tells you which races to watch live and which to record.

11 min readRead
AIMar 2026

AI copilots for companies: where the line between useful and theatre actually sits

Every company over 100 staff has been pitched a copilot. Here are the five questions that separate the deployments that ship value from the ones that survive a year on goodwill alone.

12 min readRead
DataFeb 2026

Modern data pipelines for SMEs in 2026: just use Postgres

Three years building data pipelines for UK SMEs. The lesson keeps being "just use Postgres." Here is when that breaks down and what to reach for instead.

11 min readRead
CloudFeb 2026

Azure vs AWS vs GCP in 2026: pick by what you actually do

I run workloads on all three clouds in 2026. The decision is rarely about features now. It is about pricing model, ecosystem fit, and which lock-in you can live with.

13 min readRead
AIFeb 2026

KB-grounded AI chatbots: what actually works for SMEs

I built four KB-grounded chatbots for UK SMEs in 2025. The data on ticket deflection, where users still escalated, and the cost per resolved query.

12 min readRead
CloudFeb 2026

Why I keep choosing Supabase over Firebase, Planetscale, and Neon in 2026

Three years deep on Supabase. The Postgres-native model, the free tier, and the auth integration are why I default to it for new projects. Here are the numbers and the trade-offs.

12 min readRead
AIFeb 2026

AI automation for businesses in 2026: what works, what is theatre

I have shipped AI automation for ten UK SMEs. The data shows where AI saved money and where the implementation was theatre. The pattern repeats.

14 min readRead
CloudJan 2026

Azure OpenAI vs AWS Bedrock vs GCP Vertex in 2026: where to actually run Claude, GPT, and Gemini

Three clouds. Three model marketplaces. Different pricing curves and very different lock-in stories. The data shows which one fits which workload.

14 min readRead
AutomationJan 2026

Two years of n8n: when low-code automation actually beats writing code

I have shipped 30+ workflows on n8n for clients and myself. The data on which workflows belong in n8n and which belong in code.

12 min readRead
OpinionJan 2026

My 2026 stack as a solo engineer

The exact tools I use to ship side projects and client work in 2026. Each picked for low operational cost, low monthly bill, and high leverage. Here is the list with prices.

12 min readRead
AIJan 2026

Meta's AI strategy in 2025-2026: brilliant or panicked?

Meta open-sourced Llama, hired aggressively from OpenAI, and shipped Meta AI inside WhatsApp. Are they winning or burning?

12 min readRead
AINov 2025

v0 changed how I prototype: a year of data on speed and quality

A year using v0 by Vercel for early-stage UI work. The data on time-to-prototype, percentage of v0 output that survives to production, and where it still fails.

11 min readRead
ToolsOct 2025

My v0 prototyping workflow: what to keep, what to throw away

A practical workflow for getting useful prototypes out of v0 and turning them into production code without inheriting v0 mess.

9 min readRead
ReviewsOct 2025

Five years of iPhone Pro Max: 13 to 17, charted

Weight, battery, screen, chip benchmarks and price across five iPhone Pro Max generations.

13 min readRead
ReviewsSept 2025

iPhone 17 Pro Max: thinner, but at what cost? A charted review

Apple cut 0.6mm off the iPhone Pro Max in 2025. The phone feels meaningfully better in hand. Battery life regressed. Camera bump grew. The data shows where Apple traded what.

13 min readRead
ReviewsSept 2025

Galaxy Z Fold 7 long-term: the foldable that fixed the weight problem

Eight months with the Galaxy Z Fold 7. 215 grams, an 8.0-inch inner display, and the smallest crease Samsung has shipped. The right call for the first foldable buyer in years.

11 min readRead
ReviewsAug 2025

Pixel 9 Pro Fold one year on: the lighter foldable that still trades wrong

A year with the Pixel 9 Pro Fold. Lighter than the Z Fold 6, smaller crease, better camera, weaker software ergonomics for productivity. The data shows what each side wins.

11 min readRead
CloudAug 2025

Why Supabase keeps winning for me (long version)

Supabase has been the database for every project I have shipped in the last two years. The Postgres-native model, the auth, and the realtime engine are why. Here is the long version, with numbers.

11 min readRead
F1Jul 2025

F1 2025 mid-season data: McLaren or Ferrari for the constructors

After Silverstone the WCC is wide open. Three teams within 60 points. The data tells you which one closes the deal.

9 min readRead
ToolsJun 2025

Why I switched all my domains to IONOS

Twelve domains across Namecheap, Google, GoDaddy, and 123-reg consolidated to IONOS. Cost, control panel, and DNS. The data shows where the savings came from and where I gave up small things.

8 min readRead
ReviewsMay 2025

MacBook Pro M4 Pro 14": six months in, the right Mac for most builders

Six months on the 14-inch MacBook Pro M4 Pro. The screen, the sustained performance, and the battery are the things that compound. The price is what it is.

13 min readRead
ReviewsApr 2025

MacBook Air M4 review: the laptop you should actually buy

For 95% of people, the right laptop in 2025 is the £999 MacBook Air M4. The Pro is a niche product. Here is the case.

10 min readRead
ReviewsApr 2025

Galaxy S25 Ultra long-term review: the boring upgrade that just works

Three months on the S25 Ultra. Snapdragon 8 Elite is genuinely fast. The bezels are smaller. Battery life is the best in the Ultra line. The S Pen lost Bluetooth and nobody noticed.

12 min readRead
ReviewsFeb 2025

I run my home server on a Mac mini M4 Pro: the data after three months

Three months of Caddy, Postgres, Redis, mediasoup, and a couple of LLM endpoints on a Mac mini M4 Pro. Here is what worked, what broke, and the bill.

12 min readRead
ReviewsFeb 2025

Galaxy S25 Ultra: incremental, in the right places (charted)

Year three of the new S25 Ultra design language. Smaller bezels, lighter chassis, faster chip. The data shows where it matters and where Samsung played it safe.

11 min readRead
ReviewsJan 2025

iPhone 15 Pro Max revisited: was the titanium worth it?

Two years with the iPhone 15 Pro Max. Titanium changed the feel, A17 Pro changed the thermals, USB-C changed the I/O. The data shows what stuck.

11 min readRead
F1Dec 2024

F1 2024 season: a data-driven recap

Verstappen took his fourth title but the season was tighter than the headline suggests. McLaren ran away with the Constructors. The numbers tell the story better than the commentary did.

14 min readRead
ReviewsDec 2024

iPhone 16 Pro Max long-term: the best iPhone if you can afford it

Three months with the 16 Pro Max. Camera button, A18 Pro, Apple Intelligence. The data shows where Apple actually moved versus where the marketing did.

12 min readRead
ReviewsNov 2024

Mac mini M4: the redesign Apple should have done years ago

A 12.7cm cube replaces the decade-old footprint. Same chassis to the eye, half the volume. The M4 inside is a 25% jump on M2. The data shows where Apple finally moved.

11 min readRead
ReviewsOct 2024

Apple TV 4K (3rd gen): still the best streaming box, by a long way

Two years with the third-gen Apple TV 4K. The competition has not caught up. The data shows how far ahead it is on app launch, picture quality, and update support.

8 min readRead
ReviewsOct 2024

Sonos Era 300 with Spatial Audio: the small surprise of 2024

Six months with the Sonos Era 300. The Spatial Audio rendering on a single speaker should not work. It mostly does. Here is what the actual experience is.

9 min readRead
ReviewsSept 2024

iPhone 13 Pro Max three years on: the pre-titanium peak

Three years on the 13 Pro Max. Battery health 81%. Still gets iOS updates. Still feels fast. The data shows where it has aged and where Apple Silicon kept it relevant.

10 min readRead
ReviewsSept 2024

Pixel 9 Pro Fold vs Z Fold 6: closing the gap

Two flagship foldables side by side. Samsung's lead in foldables is no longer obvious.

10 min readRead
ReviewsSept 2024

MacBook Air M3 long-term: still the default laptop for 90% of buyers

Eighteen months on the MacBook Air M3 13-inch. It does everything most people need and nothing they do not. The data shows where the limits are.

11 min readRead
ReviewsAug 2024

Galaxy Z Fold 6: the foldable that finally makes sense

Six generations in, Samsung made a foldable that does not feel like a beta. Lighter, smaller crease, brighter screens, real software for productivity. The data shows where Samsung finally caught up.

12 min readRead
F1May 2024

Norris's first F1 win at Miami 2024: a data-led look at the breakthrough

Lando Norris won his 110th Grand Prix start at Miami 2024. The data shows it was overdue.

12 min readRead
ReviewsApr 2024

Galaxy S24 Ultra three months on (archived 2024 review)

Three months with the S24 Ultra. Galaxy AI is mostly hype, the camera tuning is divisive, the chip is a meaningful jump. Archived for context against the S25 review.

10 min readRead
F1Mar 2024

F1 2024 Bahrain: Verstappen wins at a canter, the championship is over by lap 1

Bahrain 2024 was the most one-sided season opener in a decade. Verstappen took pole, the lead, fastest lap, and the win by 22 seconds.

11 min readRead
ReviewsSept 2022

iPhone 13 Pro Max one year on (the original review, archived)

My 2022 take on the iPhone 13 Pro Max after a year. The first ProMotion iPhone, the camera leap, the battery story. Archived for context against the three-year retrospective.

8 min readRead

Let's build something good.

You've got a problem. I solve problems with software for a living.The fastest way to find out if we can work together is to talk.

Get in touchTell me about your project

Stack I build with

Next.js 16TypeScriptPythonSupabasePostgreSQLVercelCloudflareDigitalOceanResendn8nClaudeOpenAITailwind v4