Mobile Apps
React Native and Expo, shipped through Apple and Google review.
Production work · fixed deadline
Broken release, urgent Telegram bot, mobile MVP, data automation or unstable infrastructure — one builder owns the path from the problem to a working production result.
React Native and Expo, shipped through Apple and Google review.
Accounts, AI assistants, image recognition, payments and access.
Proxy-backed headless collection that survives daily change.
Aggregates, coefficients and regional signals from raw data.
Store assets, slides and motion generated from live inputs.
Checks, messenger alerts, watchdogs and self-recovery.
Cloudflare, Supabase, VPS, systemd, cron and remote backups.
Source attribution without cookies or consent-banner clutter.
Task leases, multiple agents, result checks and a work journal.
Speech synthesis, voice commands and message processing.
The app will not build or Apple will not approve it.
The release is burning and the previous developer disappeared.
You need a Telegram bot for a real process — yesterday.
Data is trapped behind a login and manual work keeps growing.
Production runs on hope, one VPS and undocumented commands.
AI agents burn tokens, collide and still leave the job unfinished.
Built, configured, released and checked in the environment where it has to work.
The repository, accounts, credentials and deployment access stay under your control.
One week after delivery is reserved for corrections inside the agreed scope.
The price is fixed after the written scope. No hourly meter and no surprise expansion halfway through.
A React Native and Expo application built through the complete release path: production setup, App Store, Google Play and Apple review.
Accounting, AI assistants, image recognition, payments and access control connected to the real workflow and deployed.
Authorized login, proxies, resilient headless collection and a daily recalculation pipeline for data that cannot be exported cleanly.
Raw records turned into aggregates, operating coefficients and regional maps that make the next decision visible.
Store screenshots, social slides and motion assets assembled repeatedly from live product data instead of rebuilt by hand.
Server and service checks, messenger alerts, watchdogs and bounded self-recovery for the failures that happen after launch.
Cloudflare Workers and Pages, Supabase, VPS, systemd, cron and off-server backups set up as one maintainable production path.
Source attribution and useful web traffic reporting without cookies, invasive tracking or consent-banner clutter.
Several agents share one job through task leases, automatic result checks and a readable journal of completed work.
Speech synthesis, voice commands and voice-message processing connected to a concrete product or operating process.
Two production releases, their real store paths and the difficult parts that happened after the code looked finished.
A trip, fuel, expense and income tracker for US owner-operators. Built through the complete mobile path: offline state, sync, accounts, release builds and store submission.
React Native · Expo Router · SQLite · Supabase · EAS
Apple rejected the submission under 4.3(a) because an older version of the same product used the same icon. The duplicate was removed, the review context was explained and the build was resubmitted. Release work means handling the rejection without losing the product.
The first production version that proved the workflow with real drivers: trips, fuel, documents, income and the operating backend behind it.
Base44 · Supabase · Cloudflare · store packaging
Separate platform packaging, production access, store metadata and backend continuity had to work before the native replacement existed. The hard part was keeping a live product usable while the next version was being rebuilt.
Why there are no testimonials here: I do not turn private client messages into marketing copy. The store listings, production product and visible release history are the proof until clients explicitly approve public case studies.
You describe the problem, current state and the result that has to exist.
You receive the scope, exclusions, total price and delivery deadline in writing.
Work starts after 50%. You see working checkpoints, not ceremonial status reports.
The result is delivered with source, access and the seven-day correction window.
Because private messages are not public evidence. Until a client explicitly approves a named case study, I show products that can be opened, store releases that can be checked and the exact work completed.
You hear about a risk before the deadline, not after it. If a delay is on my side, the remaining 50% stays unpaid until the agreed scope is delivered; we then confirm the corrected date or close the undelivered part in writing.
You do. The repository, project accounts, deployment access and production credentials are handed over or created under your control. Reusable internal tooling remains mine; everything built specifically for the agreed product is yours.
50% after the written scope is approved and 50% when the agreed result is delivered. The invoice and cross-border payment method are confirmed before work starts.
One message is enough
No discovery call required. A short written description is enough to start.