1. Methodology

Slice it. Ship it. Measure it. Hand it over.

The Eaglum Method — how senior engineers modernise the systems you already run, and build the products you have not started yet, without big-bang rewrites or generic agile theatre.

The Eaglum Method

Four commitments that hold whether the engagement is a single slice of modernisation work or an end-to-end product build. They decide what counts as done — and what counts as enough.

Slice it

We cut work into shippable slices that survive contact with production. Big-bang rewrites stall — narrow, well-scoped slices keep the live product moving while the next one gets modernised.

Ship it

Each slice goes to production with tests, observability and a runbook. We treat "demo on the laptop" as half the job — the other half is the deploy, the dashboards and the on-call notes.

Measure it

We agree what success looks like before we start, and instrument it on the way out. Conversion, latency, error rate, support tickets — whichever signal matters for the slice in front of us.

Hand it over

Your team owns the code, the schemas, the infrastructure and the documentation. Knowledge transfer is part of the engagement, not a closing meeting.

Why our approach is different

Slice-based delivery is a deliberate choice — not the default. Here is how it lines up against the two patterns most teams have already lived through.

EaglumBig-bang rewriteOffshore body-shop
Starting pointAudit the running system. Cut the first slice with your engineers in the room.Spec a clean replacement of the whole product. Freeze the roadmap.Take a brief, return a quote. Begin in a parallel repository.
Team shapeSenior engineers in your timezone, embedded in your repo, your CI and your standups.Staff a project team behind a contract boundary. Demos every two weeks.A delivery floor in another timezone, brokered through an account manager.
First deliverableA shipped slice in production with tests, observability and a runbook.A migration plan, an architecture diagram, a date six months out.Status reports and burndown charts.
Risk profilePer-slice — bounded blast radius, reversible, measurable on the way out.Concentrated at cutover. Rollback is rarely possible without losing months.Spread across the contract — quality drift caught at handover, not on the way in.
At handoverCode, infrastructure, runbooks and a team that has paired with yours throughout.A new system your team has not touched and now has to operate.A repository with patterns your engineers did not choose and may not recognise.

A slice, end to end

Every slice runs through the same five phases. The ranges below are honest — production rarely lands on a point estimate, and we would rather agree the range up front than miss it later.

01

Discovery

1–2 weeks

Read the codebase, the dashboards and the on-call history. Talk to the engineers and the people closest to the users. Map the system as it actually runs — not as the documentation describes it.

02

Slice definition

1 week

Agree the first slice with your engineers. Write the Slice Definition Doc — scope, success metric, Definition of Done, risks. If we cannot defend it on a page, it is not the right slice.

03

Build the slice

4–8 weeks

Pair, review, test, deploy. The slice goes to production behind a feature flag with the observability and runbook the team will need on the next quiet weekend.

04

Harden and measure

1–2 weeks

Watch the metric we agreed. Fix the things production teaches us. Tighten the runbook with the alerts that actually fired, not the ones we expected to fire.

05

Handover or next slice

1 week

Walk your team through the code, the infrastructure and the runbook. Decide together whether to scope the next slice or step back. The choice is yours, with no contract pressure either way.

The Playbook

Ceremonies and artefacts that show up on every engagement. Each one has a job to do — none of them exists because the framework asks for it.

Kickoff Workshop

A two-hour working session at the start of every engagement. Aligns engineers, designers and stakeholders on the slice, the success metric and the Definition of Done — written down, shared, and revisited.

Slice Definition Doc

A single page covering scope, success metric, risks, Definition of Done and roll-back plan. The contract between Eaglum and your team for the slice in front of us.

Architecture Walkthrough

A working session — not a slide deck — through the part of the system the slice will touch. We surface assumptions before they cost us a sprint.

Definition of Done

Tests, observability, runbook, documentation, security review. A slice is not done because the demo works — it is done when the next on-call engineer can look after it.

Runbook

Every shipped slice leaves with a runbook covering the alerts that fire, what they mean and what to do about them. Written by the engineers who built it, for the engineers who will inherit it.

Handover Pack

Code, schemas, infrastructure-as-code, runbooks, an architecture overview and a recorded walkthrough. Enough for your team to keep building without us in the room.

Sprint Demo

Working software, every two weeks, in the environment it will run in. We demo the slice and the metric — not slides about the slice.

Post-launch Review

Two to four weeks after the slice ships, we review what production taught us. Findings feed the next slice or the handover pack — never just an internal retro.

How we engineer

The commitments that decide whether a slice is shippable — not just buildable. We hold to these whether the work is greenfield, a multi-version Laravel upgrade, or a backend that needs to come off vanilla PHP.

Code review with no merge alone

Every change goes through a pull request and a second engineer. No solo merges to main, no "small fix" exceptions. The review is part of how the work gets done, not a tax at the end.

Pairing where it pays

New engineers, sensitive code paths, gnarly debugging — pairing earns its cost. We pair when the work warrants it, not as a ritual.

Tests at the layer that matters

Pest, PHPUnit or Vitest at the unit and feature layer; end-to-end where the slice spans the full stack. We do not chase coverage numbers — we cover the behaviour that production cares about.

Observability from the first deploy

Structured logs, traces and dashboards land with the first slice. When something behaves unexpectedly in production, we answer "why" without paging an engineer.

CI from day one

Pint, PHPStan, ESLint and TypeScript-strict run on every change, on a clean machine. Static analysis catches the regressions humans miss — and the cost is paid once, not per release.

Senior engineers, no follow-the-sun handover

Engineers who write the code are the same ones in your standups, working hours that overlap fully with yours. No separate delivery floor, no parallel vendor Slack, no late-night context drops between shifts.

How we work with your team

Four engagement shapes — picked in the first conversation, not assumed up front. The method holds across all of them.

Slice-based modernisation

For teams with a working system that needs to move forward without freezing. We scope a single slice, ship it under live traffic, and decide together what comes next.

Greenfield product builds

For teams with a problem worth solving and the conviction to commit. One team across design and engineering, shipping in production from week one rather than the demo at week twelve.

Embedded engineers

Senior engineers join your existing team, in your repo, your CI and your standups. They follow your conventions, not ours. When the engagement ends, your team has the code and the institutional knowledge.

Two-week Slice Discovery

A fixed-scope, fixed-price audit of a system you already run — with the first slice scoped, sized and recommended. The engagement that decides whether to hire us for the build, with no contract pressure if the answer is "not yet".

What you take home

Handover is part of the engagement, not an afterthought at the end. Every Eaglum engagement leaves your team with everything they need to operate, extend and own the slice without us.

  • All source code, on day one — under your organisation, on your version control, with no licensing strings.
  • Database schemas, migrations and seed data, with the production-shape examples your team will need.
  • Infrastructure-as-code for every environment we touched — Forge, Vapor, Terraform, GitHub Actions.
  • Runbooks for the alerts that actually fire in production, written by the engineers who built the slice.
  • A documented architecture overview that matches the running system — not the version we wished we had built.
  • A recorded walkthrough of the slice, the rollout, and what to watch in the first month.

Frequently asked questions

How is this different from generic agile?

Generic agile is a meeting cadence. Our method is a set of commitments — slice the work small enough that each piece ships, treat shipping as deployed-and-measured rather than demoed, and hand the work over with the runbooks and documentation your team will need on a quiet Sunday. The ceremonies serve those commitments, not the other way around.

What happens if a slice runs long?

We surface it the moment we see it, not at the end. The Slice Definition Doc names the risks at the start, so a slice running long usually traces back to a known risk — and we re-scope, ship a smaller slice, or bring in a second engineer, depending on what the risk is. We do not absorb scope creep silently.

Do you work in our repos and our CI, or your own?

Yours. Engineers join your repo, your CI, your review process and your release cadence on day one. We do not run a parallel project in a separate workspace — the work happens where your team will eventually own it.

Who writes the runbooks and the documentation?

The engineers who built the slice, while the work is fresh. Documentation written by a separate person at the end of the engagement is documentation written from memory, and it shows. We treat runbooks as part of Definition of Done, not part of handover.

Can we hire your engineers permanently?

On a case-by-case basis. We are open to discussing it when the engagement is the right fit for everyone — including the engineer. The goal of every engagement is the best outcome for your product, which is sometimes us staying and sometimes us stepping back.

What do we own at handover?

Everything. Source code, schemas, migrations, infrastructure-as-code, tests, runbooks, documentation, recorded walkthroughs. Under your organisation, on your version control, from day one. No proprietary tooling, no licensed components, no lock-in.

How fast can we start?

Most engagements begin within one to two weeks of the initial conversation. Two-week Slice Discovery usually starts faster — sometimes the same week — because the scope is fixed.

Do you take on fixed-price work?

For fixed-scope engagements like Slice Discovery, yes. For build engagements we work time-and-materials with a clearly defined slice and a Definition of Done — a fixed price on an open scope is a fiction nobody benefits from.

Tell us about the slice in front of you

A two-week Slice Discovery turns a system you already run into a scoped first slice — with a roadmap, an estimate and no contract pressure.

Send an email