Available for new AI & automation projects

Your business bottlenecks,
turned into deployed AI systems.

You don't need another coder waiting for a perfect spec, or a consultant who hands you a slide deck. You need someone who understands the business and builds the system - the rare overlap. After 15+ years running P&Ls, fundraising and go-to-market, I now build the AI myself, so the bottleneck costing you time and money actually gets found and fixed.

Your real bottleneck, foundDiagnosed like a business developer - because I've run the P&L, not just the codebase.
Time and money backHours of manual work collapse into seconds - one builder, not an agency.
Live in productionWorking systems you own - not slide decks or prototypes.
15+ yrs
Founding & scaling tech companies
$15M+
Capital raised across ventures
13+
AI systems built & shipped
6
LLMs in production pipelines
Why work with me

A business builder who
also writes the code.

Most AI projects die in the gap between the boardroom and the codebase. I've lived on both sides of it.

I spent 15+ years as a founder, CEO, GM and VP - raising capital, running multinational projects, taking products from idea to market and to exit. Then I went deep into AI and learned to build the systems myself.

So when we talk, I understand your P&L, your customer and your go-to-market - and because an operational bottleneck looks the same whether you're in finance, healthcare or supply chain, I can architect the pipeline and push it to production the same week. No translation layer between what the business needs and what gets built.

Founder & operator

Founded & exited an AI industrial-safety startup; co-founded a medtech venture, raised $6M+, tripled valuation through Series A.

Business development & GM

GM / VP roles in medtech, agtech & digital health - led $3–5M projects, fundraising rounds, and US / EU / Asia market entry.

Deep technical foundation

MSc Chemical Eng., BSc Materials Eng., BA Physics - Technion. An engineer's rigor behind every system.

Ships to production, solo

13+ live AI systems: multi-agent pipelines, LLM automation, real-time dashboards, data engines - all deployed.

What I do

AI integration & automation, end to end.

Whether you need one workflow automated or a full intelligent platform - I scope it, build it, and deploy it. Here's where I create the most value.

Multi-Agent AI Systems

Orchestrated pipelines where specialized AI agents collaborate on complex workflows - lead handling, analysis, decisioning - delivering in seconds what took hours.

Workflow Automation

Automate the repetitive, error-prone work: document processing, data entry, outreach, reporting. Reliable systems that run 24/7 without you touching them.

LLM Integration

Add GPT, Claude, Gemini or Grok to your product or operation the right way - RAG, vision, function-calling, the correct model for each job, cost-controlled.

Data & Scraping Pipelines

Extract, clean and structure data from anywhere - web, PDFs, APIs, competitor catalogs or public records - and pipe it straight into your database or dashboard.

Dashboards & Real-Time Apps

Live, decision-ready dashboards - WebSocket streaming, interactive charts, sub-second data. See what matters the moment it happens.

Deploy & Own It

Dockerized, on your cloud or mine, with SSL, monitoring and CI/CD. You get a working system in production - not a prototype in a repo.

Selected work

Real problems, shipped as working systems.

A selection of live, production platforms. Every one started as someone's real bottleneck.

LeadKong AI CRM
Multi-Agent SaaS · Live

LeadKong - 9-Agent AI Deal Engine

The problem: a flood of unstructured inbound data across 8+ channels, where analyzing each item by hand took hours - so warm opportunities went cold. Nine specialized AI agents now ingest, structure and analyze complex visual and textual data in under 60 seconds each, then surface a ranked shortlist with follow-up messages drafted and ready. Built for real-estate deal flow - but the same pipeline fits any business with high-volume intake to triage.

Next.jsFastAPIGrok · GeminiPostgreSQLGoogle Cloud Run
Live at leadkong.net
MirageMend marketing automation
AI Marketing Automation · Live

MirageMend - Autonomous Outreach Platform

The problem: generic mass email gets ignored, but genuinely personalized outreach doesn't scale to thousands of recipients. MirageMend uses AI to write context-aware messages for each individual recipient, runs multi-stage drip sequences across rotating inboxes, and detects replies over IMAP - advancing every lead through a Kanban CRM automatically. Outreach that reads one-to-one, at a volume no human team could sustain - whatever you're selling.

FlaskGrok · GeminiSMTP / IMAPAPSchedulerOracle Cloud
Live at miragemendrealty.com
Private client projectVisuals under NDA
Full-Stack Business App · Private client · Live

Proposal & Quote Management Platform

The problem: a services business ran its whole sales pipeline out of spreadsheets, email and hand-built PDFs - with no view of which offers were sent, approved or stalled. I built a Kanban proposal platform that carries every offer from draft to completed, auto-generates polished PDFs into the right Google Drive folders, and tracks milestones and payments. Full Hebrew RTL, Google OAuth, dockerized behind SSL - issuing an offer went from an afternoon to two minutes.

React · TSFastAPISupabaseDocker · NginxGoogle Drive API
Real-time community alert pushed to WhatsApp
Real-Time Alerts · Live 24/7

Community Live-Alert Bot

The problem: small communities need life-safety alerts - like Home Front Command rocket warnings - to reach every resident the instant they fire, but official apps are easy to miss and no one checks them in time. This bot ingests the alert API in real time and pushes each warning straight into the WhatsApp groups people already watch, 24/7, with zero human in the loop. When seconds matter, the warning is already on everyone's phone - and the same event-to-WhatsApp pipeline fits any critical notification: outages, on-call escalations, logistics events.

PythonWhatsApp APIReal-time alert API24/7 daemonOracle Cloud

…and a deep bench of automation & AI tools

TaseFuturesWithAI

4 LLMs in parallel generating trading predictions across 30+ indicators. ~40K lines of Python.

TASE Real-Time Monitor

Live options terminal - SignalR streaming, in-browser Black-Scholes, sub-second latency.

BRRR Calculator Pro

Hybrid AI + scraping property analyzer with ROI modeling and Google Maps. Cloud-deployed.

AI Property Monitor

Autonomous agent that screens listing photos with Claude Vision against custom criteria 24/7.

Court Records Pipeline

12-stage scraper extracting property data from court records, analyzed with AI, auto-organized.

Geospatial Vision Tool

OpenCV analysis of map imagery - detects properties & water, ranks by proximity.

Stay current

Latest in AI

A live feed of what's happening in AI - auto-updated so you never fall behind.

More news
FAQ

Questions, answered.

The things people usually ask before we talk.

What does an AI integrator actually do?
I find where your business loses time or money, then design, build and deploy a working system - using AI where it genuinely helps - to remove that bottleneck. One person owns the whole thing: AI design, backend, frontend, cloud and deployment. No handoffs, no gaps.
How is this different from an agency or a freelance developer?
Most developers can code but don't understand your business; most consultants understand the business but can't build. I'm the rare overlap - a founder who ran P&Ls, fundraising and go-to-market for 15+ years, and now builds the AI myself. You get one accountable person, faster delivery, and no agency markup.
Which industries do you work with?
Any of them. An operational bottleneck looks the same whether you're in finance, healthcare, logistics, e-commerce or manufacturing. My past work spans real estate, trading, medtech, agtech and industrial safety - and the underlying mechanisms (structuring messy data, personalizing communication at scale, automating workflows) transfer directly to your business.
How quickly can you deliver?
Fast. We usually start with a scoped pilot that ships to production in days to a few weeks, then iterate from there. The goal is always a working system in production, not a prototype in a repo.
How do we start, and what does it cost?
It starts with a free call to understand the problem - and whether AI is even the right fix. From there you get a clear scope and price for a first deliverable, so there's no open-ended commitment. Use the contact form to get started.
Let's build something

Have a workflow that
should run itself?

Tell me what's slowing you down. I'll tell you honestly whether AI is the right fix - and if it is, how I'd build it. No jargon, no obligation.

Goes straight to my inbox - I usually reply within a day.