[ 00 / 07 ] · HOME
accepting Q3 projectsmaalot-tarshiha · remote-first

We build software
that actually ships
— and the cloud it runs on.

Phoned-Tech is a senior engineering studio. We deliver production web, mobile, and AI systems — from first commit to deployed infrastructure. Including our own SaaS products, with the servers, cloud, and domain set up for you.

11+years shipping
40+production systems
24/7infra monitored
#1Quality
~/phoned-tech/intro.shzsh · 80×24
$ whoami
phoned-tech · senior software studio · since 2013
 
$ ls ./capabilities
▸ web        
react · next · vue · typescript
▸ mobile     
ios · android · react native
▸ backend    
node · laravel · spring boot · php
▸ ai         
rag · agents · fine-tuning · evals
▸ devops     
aws · gcp · k8s · docker · ci/cd
▸ saas       
phoned-crm · servers · domains · cloud
 
$ deploy --status
build  
✓ passing
tests  
✓ 1,284 passing · 0 failing
deploy 
✓ live · 4 regions · 99.99% uptime
 
$ echo $READY_FOR
your next project →

We also do AI — and we do it seriously.

AI services
RAG01

Retrieval-augmented generation

Chunk, embed, retrieve. Index your docs, code, tickets — and get grounded answers with citations.

  • Hybrid search (vector + keyword)
  • Re-ranking, query rewriting
  • Source citations + audit trail
Agents02

Tool-using AI agents

Agents that call your APIs, write to your DB, and ask for human approval when stakes are high.

  • Tool definitions & schemas
  • Memory, planning, retries
  • Human-in-the-loop checkpoints
Fine-tuning03

Fine-tuning & evals

When prompting isn't enough. Curate data, fine-tune, run evals — measure before you ship.

  • Dataset curation & labeling
  • SFT / DPO / LoRA pipelines
  • Eval suites & regression tests
Automation04

Workflow automation

n8n, Make, custom orchestrators. Replace the spreadsheet, the inbox, the copy-paste.

  • Triggers, queues, retries
  • Vendor failover
  • Cost & latency dashboards
▸ /llms.txt — site is LLM-ready

Designed for humans and machines.

Semantic HTML, JSON-LD, sitemap, OpenGraph, and an /llms.txt manifest so retrieval engines and AI assistants index every page accurately.

GET /llms.txt
> 200 OK
> Phoned-Tech
> Software · AI · SaaS
> 7 pages indexed
> last-modified: 2026-05-05

Boring tech, where it counts.

Full stack

// frontend

  • React / Next.js
  • Vue / Nuxt
  • TypeScript
  • React Native

// backend

  • Node.js
  • Laravel · PHP
  • Java · Spring Boot
  • Python

// data

  • PostgreSQL
  • Supabase
  • MongoDB
  • Redis

// cloud & ai

  • AWS
  • Google Cloud
  • Vercel · Cloudflare
  • Docker · Kubernetes

How we work.

5 phases · 4–16 weeks typical
  1. 01

    Discovery

    Free 60-min call. We map your domain, constraints, and success metrics. You leave with a written one-pager — even if you don't hire us.

    ~1 week
  2. 02

    Architecture

    Stack choices, data model, infrastructure plan. Documented decisions you can hand to any engineer in five years.

    1–2 weeks
  3. 03

    Build

    Two-week sprints, demo every Friday. Production deploys from day one. You get the repo, we get the discipline.

    2–12 weeks
  4. 04

    Launch

    DNS, SSL, monitoring, runbooks. We set up your cloud, your domain, your alerts. You wake up to dashboards, not pages.

    ~1 week
  5. 05

    Operate

    Optional retainer. We keep it running, patched, observed, and improving. SLAs available.

    ongoing

A CRM that comes with the cloud.

Phoned-CRM is our flagship SaaS — a customer platform you can adopt without an IT department. We provision your server, cloud, domain, SSL, mailers, and backups. You log in to a working system.

  • Multi-tenant or dedicated server
  • Custom domain & branded login
  • Pipelines, automations, integrations
  • AI assistant included (RAG over your data)
  • Backups, monitoring, 24/7 infra
app.your-company.com / phoned-crm
Pipeline
Contacts
Inbox
Automations
AI Assistant
Reports
Lead · 12
Acme Holdings
$24,000
Northwind
$8,400
Globex
$52,000
Qualified · 7
Soylent Corp
$96,000
Initech
$12,000
Won · 3
Hooli
$140,000

Things people always ask.

How long does a typical project take?+

Anywhere from 4 weeks for a focused MVP to 16+ weeks for a full SaaS platform with infrastructure. Discovery is always the first week — we won't quote without it.

Do you work fixed-price or time & materials?+

Both. We default to two-week sprint billing with a written scope per sprint. Fixed-price works well for tightly-scoped deliverables — landing sites, integrations, AI prototypes.

Who owns the code and infrastructure?+

You do. From day one. Repo on your GitHub org, cloud account in your name, domain registered to your company. We just have keys.

Can you set up our cloud, domain, and servers from scratch?+

Yes — that's standard for our SaaS deployments and a common add-on for everything else. We provision AWS or GCP, register domains, configure DNS/SSL, mailers, monitoring, and backups.

Do you replace our team or work alongside it?+

Both. We embed inside engineering teams, run autonomous squads, or operate as fractional CTOs. We're explicit about which model we're in from week one.

What about AI specifically — what should we expect?+

Honest scoping. Most "AI features" are 80% data plumbing and UX, 20% model. We tell you when prompting is enough and when fine-tuning earns its keep — backed by evals you can rerun.

[ END / 09 ]▸ /contact

Have something specific
in mind?

Book a free 30-minute call. We'll listen, sketch, and send you a written one-pager — whether or not we end up working together.