Agent harnesses, tokenomics, and team training · For the people who own the AI budget, and the teams who use it · Based in MENA, working globally

I build AI systems. I train teams to run them.

Real-time voice agents at CME Offshore and Siren Analytics. Enterprise AI systems inside Strategy& / PwC, where I work today. Alongside that, I work with companies directly: getting AI past the demo stage, with the economics done before the build.

Start with a call Choose your track
Rony, full stack AI engineer
Vero AIReal-time voice agents, built at CME Offshore and Siren Analytics alongside a great team
Strategy& / PwCAI systems for enterprise clients. Specifics under NDA
Viral AlchemyMy own build: a system that maps what is about to trend

Most AI initiatives stall between the demo and production. The gap is engineering and economics, not model choice. That gap is where I work.

A stalled pilot keeps billing every month it sits there. Nobody logs that as a failure. It gets logged as a deprioritization. The budget is gone either way.

Choose your track

Two tracks. Same standard: it works, and you know what it costs.

Both run inside your company, on your stack and your real work. Pick the team that is stuck. You can switch any time.

01 For your engineering team Build AI systems that survive production. Agent harnesses. Context engineering. Tokenomics. PMs, tech leads, juniors, seniors, frontend: everyone in the same room. Enter the engineering track 02 For your business teams Real productivity with AI, not prompt tricks. Automations. Workflows. Output quality. For operations, marketing, finance, and support. No code required. Enter the business track

Working solo? A solopreneur gets the same standard, one on one. Pick the track closest to your work and book the same call.

How it starts

One call. Then you decide.

Most teams are not choosing between me and another vendor. They are choosing between doing this and doing nothing. Doing nothing has a price too. The other options are a hire you cannot fill for months, or a consultancy that leaves a deck. Neither leaves capability behind.

Step 1

A 20-minute call

You describe where AI is stuck in your team. If I am not the right person, I say so and point you toward what would actually help.

Step 2

The Tokenomics Audit (an AI spend review)

A paid, fixed-scope review of your AI spend and pipeline, line by line. It ships as an interactive report: costs mapped, fixes ranked. You keep it whether or not we go further.

Priced on the call, scoped in writing before anything starts.

Step 3

Build and train together

We fix the system with your team in the room, so the capability stays when I leave. Your team's time is the real cost. Both numbers go on the table before we start. The goal is that you stop needing me.

Start with a call

The work

What I have built. Stated plainly.

CME Offshore · Siren AnalyticsVero AI

Real-time voice agents, live in production.

A customer-operations platform where callers reach a voice agent on phone, WhatsApp, and email, in five languages. I made the full voice chain work: SIP trunking through Twilio, LiveKit and WebRTC, speech and reasoning on both local and API models, multimodal RAG, database access as agent tools. First on the team to get the agent speaking end to end.

veroai.com

Strategy& · PwCEnterprise AI systems

AI systems for enterprise clients. Specifics under NDA.

Internal builds and demos across client engagements, several adopted into active team use. The work is real. The details stay covered until clients say otherwise.

Personal buildViral Alchemy

A system that maps what is about to trend.

My own product: creators onboard, point it at their niche, and it scans Reddit, Instagram, TikTok, and YouTube for what is rising, then turns the strongest ideas into scripts.

viralalchemy.io

The full portfolio: 15+ projects, documented and ranked

About

Rony Raffoul. Full stack AI engineer.

The stack

Python FastAPI React Next.js LLM agents Agent harnesses Custom MCPs Claude Code Multimodal RAG Real-time voice Real-time avatars LiveKit · WebRTC Twilio · SIP Docker Azure Tokenomics

Real-time voice agents at CME Offshore and Siren Analytics. That work became Vero AI. Today, enterprise AI systems inside Strategy& / PwC, alongside direct work with companies. Real-time avatars as separate hands-on work. Based in MENA.

Self-taught, starting in 2020, in a room with a laptop and no GPU. The first build: an exercise-recognition system, data filmed at a real gym, trained by hand, demonstrated live in front of the jury.

Before this there was The Learning Engineer, my brand teaching students how to learn. Neuroscience, teaching, and AI systems. The combination is rare, and it is why the training half of this practice exists.

What I will not do

  • Sell you an AI strategy deck without touching the system.
  • Take an engagement I cannot personally deliver.
  • Propose a long program when a short project solves it.
  • Quote numbers I cannot back. On this site or in a meeting.

The first step

Twenty minutes. You describe where AI is stuck.

You leave the call knowing whether I can help, roughly what it would take, and what it would cost to find out for sure. If I am not the right person, I will say so on the call.

The 20 minutes, in order

  1. You talk first. Where AI is stuck, what has been tried, what it costs today.
  2. I answer plainly. Whether I can help, and what I would look at first.
  3. You leave with a next step. An audit scope, a pointer elsewhere, or a clean no.

Before you book

  1. Your stack is not too messy to start. Messy is the normal starting condition.
  2. Training that does not survive the week is the normal outcome. These sessions build on your backlog and ship something real.
  3. You do not need budget approved to take the call. You need it approved to start step two.
Book the 20 minutes Not ready. Send me the newsletter

The build log

What I am building, what it costs to run, and what broke on the way. Monthly, at minimum. Leaving takes one click.

[Email provider pending: Kit]