davthecoderdavthecoder

For product teams · mobile and applied AI

Reduce legacy drag. Ship AI that holds up in production.

I help product teams modernise Android, adopt Kotlin Multiplatform, and ship contained AI features with senior hands-on delivery. Mobile and applied AI are both hireable tracks, not a rebrand away from either. The work begins with the smallest change that can prove the direction.

Public evidence includes employment products used by millions and independent systems such as paglipat.com and the live chat on this site. Official platform docs for Android and Kotlin Multiplatform remain the baseline for language and tooling claims.

Discuss your projectApplied AI enquiryView case studies

What mobile problems do you solve?

Ageing Android codebases that are slow to change, Android and iOS logic written twice, and teams that need senior hands-on capacity without pausing the roadmap. Typical work is Kotlin and Compose modernisation, incremental Kotlin Multiplatform adoption, and architecture the permanent team can keep running.

What is applied AI here?

Contained product features such as assistants, tool calling, grounding, guardrails, cost controls, and tests. Evidence is from independent products (Paglipat, this site’s chat, ElTarot), not invented client AI work. Not model training or MLOps as a service.

Core services

Start with the mobile constraint

Android modernisation and KMP adoption are the main offers. Embedded delivery is how I can join when the work needs ongoing senior ownership.

Core service

Android app modernisation

Problem
An ageing Android codebase is slow to change, hard to test, and expensive to keep alive.
Good fit
Legacy Java or XML, fragile architecture, slow releases, or rising crash and regression rates.
Deliverable
A prioritised migration to Kotlin and Jetpack Compose, modular architecture, practical test and CI/CD foundations, and a plan the team can continue.
Discuss this service →

Core service

Kotlin Multiplatform adoption

Problem
Android and iOS drift apart because the same business logic is written and fixed twice.
Good fit
A product team wants a shared core without pausing its roadmap or rewriting both apps.
Deliverable
An incremental migration proven on one feature, a shared module structure, native integration boundaries, and a team able to keep going.
Discuss this service →

Core engagement

Mobile architecture and embedded senior delivery

Problem
A mobile team needs senior capacity and architecture decisions that hold while the roadmap keeps moving.
Good fit
A team that needs hands-on delivery, clearer module boundaries, stronger engineering standards, or technical leadership now.
Deliverable
Hands-on senior delivery, architecture decisions, documentation, and knowledge transfer that leaves the permanent team stronger.
Discuss this service →

Specialist capabilities

Applied AI and native performance when the product needs them

Hire these as focused workstreams or as part of a wider engagement. Not broad transformation claims.

Specialist capability

Native and Rust performance work

A latency or CPU-bound mobile problem needs one reliable implementation across Android and iOS.

What I deliver: A focused Rust or native module with clean Kotlin and Swift APIs, tests, benchmarks, and documented ownership boundaries.

Specialist capability

Applied AI product features

An assistant or generation feature demos well but lacks grounding, guardrails, cost controls, and reliable tests.

What I deliver: Tool calling, grounding, model routing, prompt and tool tests, budget controls, rate limits, and retention behavior appropriate to the feature.

Evidence comes from products I shipped myself, not client AI work: the Paglipat tool calling concierge on paglipat.com, this site’s floating chatbot as a live demo, and a contained ElTarot reading feature.

Paglipat production concierge →This site’s AI assistant case study →ElTarot contained AI feature →

How I work

Clear scope, hands-on delivery, clean handover

  1. 01

    Scope

    Map the codebase, constraints, risks, and the smallest valuable first move.

  2. 02

    Deliver

    Work hands-on inside a fixed scope or alongside the team as an embedded senior engineer.

  3. 03

    Hand over

    Leave decisions, tests, documentation, and enough context for the team to own the result.

Proof

Three representative cases

Mobile flagship delivery, production applied AI, and shared native systems. Each case starts with contribution, then engineering context.

Led mobile engineering for Sky Go, combining hands-on delivery with ownership of team, architecture, modernisation, and release outcomes.

Sky Go

Engineering leadership on Sky Go across native mobile streaming, DRM-protected playback, offline downloads, and delivery for millions of subscribers.

Read case study →

Designed and built the multilingual platform, edge architecture, travel integrations, and production AI concierge.

Paglipat

An independent travel platform across 11 languages with edge delivery, a grounded tool-calling AI concierge, and an Android app on Google Play.

Read case study →

Designed and built the shared Kotlin API, native platform backends, C++ DSP layer, publishing workflow, and documentation.

Klarinet

An open-source Kotlin Multiplatform audio SDK with native Android and Apple backends plus a shared C++ DSP core for real-time effects.

Read case study →

Bring the codebase, constraint, and deadline.

I will tell you whether I am a fit and what the smallest useful first step looks like.

Discuss your project

FAQ

Common questions

Do you work remotely?

Yes. I work remotely from London with European and US-East overlap. On-site visits can be arranged for kick-off or key milestones.

Do you take fixed-scope projects or ongoing work?

Both. Modernisation and KMP adoption often begin with a bounded discovery or pilot. Embedded delivery works better when the team needs ongoing senior capacity. Applied AI work is usually a fixed-scope feature or a clear first vertical slice.

When does native Rust make sense?

When measurement shows a latency- or CPU-bound problem and one implementation across Android and iOS reduces risk. If a native layer does not earn its complexity, I will say so.

What does applied AI mean here?

Contained product features such as tool calling, grounding, guardrails, cost controls, and evaluation. Evidence comes from products I shipped myself, not client AI work: the Paglipat production concierge on paglipat.com, the floating chat on this site as a live demo, and a smaller ElTarot reading feature. I do not present model training or MLOps as a service.

What mobile problems do you solve?

Ageing Android codebases that are slow to change, Android and iOS logic written twice, and product teams that need senior hands-on capacity without pausing the roadmap. Typical outcomes are Kotlin and Compose modernisation, incremental KMP adoption, and cleaner architecture the team can own.