// our process

Five stages. Nothing hidden.

How a project actually runs at Ligiotec — from the first discovery call to the fifth year of operation. Each stage below is openly documented: what you'll get, what we'll do, and the artefacts you'll receive along the way.

12–18
Weeks · typical project
0
Missed milestones, 2024
5
Sign-off gates
Weekly
Demos · every Friday
// typical timeline

A typical 14-week project.

Stage durations overlap slightly — design ends as build begins, launch prep starts mid-build. This is the shape for a mid-sized SaaS; larger enterprise work stretches every stage proportionally.

01Discovery
1–2 weeks
02Strategy
1 week
03Design
2–3 weeks
04Build
4–12 weeks
05Launch & Grow
Ongoing
01
Listening before writing

Discovery

1–2 weeks

Two weeks of listening before a single line of code. Not because we want to pad a proposal — because the next 12 weeks depend on getting this right.

What we do
  • /01Stakeholder interviews across every role that will touch the product
  • /02Existing-system audit — what survives, what retires
  • /03Risk register — legal, technical, commercial
  • /04Competitive landscape scan
  • /05User-journey mapping with your actual users where possible
What you get
  • Brief document (8-14 pages)
  • Risk register with mitigations
  • Recommended scope (including what we recommend against)
  • Fixed estimate + milestone plan for stage 2
Artefacts you'll actually receive
Brief.pdfWritten doc. You sign off.
Risks.xlsxRanked by severity + likelihood.
Journey mapsFigjam board per persona.
Principles in this stage
If we recommend against

We've said "you don't need us" 11 times in 2024. That's the discovery working.

Named owners

Every decision has a person accountable on both sides. No "the team decided".

02
Commit before building

Strategy

1 week

Architecture, cost, timeline — committed in writing before a single line of code. The week that makes or breaks the next three months.

What we do
  • /01System architecture design (diagrams + ADRs)
  • /02Tech-stack choice with honest tradeoffs in writing
  • /03Milestone plan with deliverable gates
  • /04Fixed estimate — no "time and material" surprises
  • /05Integration inventory (APIs, SDKs, third-parties)
What you get
  • Blueprint document with architecture diagrams
  • Tech-stack rationale memo
  • Milestone plan with fixed dates
  • Fixed cost + change-order policy
Artefacts you'll actually receive
Blueprint.figVisual architecture.
ADRs/*.mdOne decision per file.
Milestones.csv6-14 dated gates.
Principles in this stage
Boring stacks

We pick tech that will still be maintained in 2030, not tech that's hot this quarter.

Fixed is fixed

We honour fixed estimates even when they hurt us. It forces better estimation.

03
Prototypes you can click

Design

2–3 weeks

Figma prototypes, user-tested with real users where possible. You sign off before we open a code editor.

What we do
  • /01Design system foundations (tokens, components, patterns)
  • /02Information architecture + user flows
  • /03High-fidelity screens for every primary journey
  • /04Interactive prototype with real transitions
  • /05Usability test with 3-5 real users + report
What you get
  • Figma file (system + screens + prototype)
  • Design system documentation
  • User-test findings + priority fixes applied
  • Assets pack (SVG exports, Lottie specs)
Artefacts you'll actually receive
v2.fig187 frames. Reviewed live.
DesignSystem.mdToken + component docs.
UserTest.pdf5 sessions transcribed.
Principles in this stage
Sign-off is sacred

Once you sign, we ship. Scope changes go through a written change-order — never a hallway conversation.

Test before shipping

Five real users catch 85% of usability issues. We don't skip this.

04
Two-week sprints. Friday demos.

Build

4–12 weeks

Engineering kicks in. Two-week sprints, live demos every Friday, no 12-week "trust us" periods. You see it work or you see what broke.

What we do
  • /01Sprint planning every two weeks with you in the room
  • /02Daily code review — every PR reviewed by a senior
  • /03Automated QA runs on every merge
  • /04Weekly Friday demo call (recorded)
  • /05Fortnightly retrospective — you co-write the notes
What you get
  • Working software at every Friday demo
  • Sprint burndown + velocity charts
  • Full CI/CD pipeline with staging
  • Automated test coverage 80%+ on critical paths
Artefacts you'll actually receive
PR #247 mergedSquash + merge + deploy.
demo-week-08.mp4Recording you can rewatch.
Retro.mdWritten after every sprint.
Principles in this stage
Weekly, not monthly

A project that can't demo weekly is a project hiding problems. Ours can.

Senior code review

Every line ships through a senior engineer. Juniors learn. Bugs don't leak.

05
After the confetti

Launch & Grow

Ongoing

Most agencies vanish after launch. We don't. Deploy, monitor, iterate. 84% of our clients stay past year one; 60% past year three.

What we do
  • /01Staged production rollout (5% → 25% → 100%)
  • /02Observability setup — alerts, dashboards, on-call
  • /03Runbook + incident-response playbook
  • /04Monthly health report to you
  • /05Quarterly roadmap review
What you get
  • Live production system with SLOs
  • On-call rotation (if in scope)
  • Monthly metrics deck
  • Quarterly roadmap artefact
Artefacts you'll actually receive
p95 180ms · 99.97%Monitored continuously
Runbook.mdEvery common failure.
MonthlyReport.pdfOne page. Real numbers.
Principles in this stage
Zero missed deadlines

Since 2022. It's a public number and we protect it.

On-call stays ours

For 90 days post-launch, then optional retainer. If it paged at 3 AM, we reply.

// questions we get

Questions founders actually ask.

/01Can we skip Discovery if we already have a PRD?

We do a shortened version — usually 3-5 days — reviewing your existing docs and running 2-3 stakeholder calls. We never skip it entirely because the assumptions that get baked in during that week are expensive to unwind later.

/02What if scope changes mid-build?

You write a change order, we estimate impact, you approve in writing. No hallway decisions, no "it's a small change" accumulation. This is the single most important discipline for staying on time and budget.

/03Do we really get weekly demos even in early weeks?

Yes. In week 1 the demo might be an architecture walkthrough; in week 3 a design review; from week 5 onward, working software you can click. The ritual is the point — it forces us to have something worth showing.

/04What happens if you miss a milestone?

We haven't missed one since 2022. If we ever did, the change-order policy works in reverse — we absorb the cost of the overrun because the fixed estimate was ours to make.

/05Who from our side needs to be involved?

Minimum: one decision-maker who can sign off, available for a weekly 30-minute demo. Ideally: a product owner for day-to-day questions (2-3 hours/week). We don't need more than that.

/06Do you work with clients outside India?

Yes. We're structured for asynchronous + timezone-overlap work. UK / EU / US-East / AUS clients get 4-6 hour IST overlap and async written handoffs the rest of the day.

// ready to start?

Every project we've done starts the same way.

One 30-minute discovery call. No slide deck. You leave with three specific next steps — whether or not we work together.