Distributed Systems & AI Engineering Advisor

Fix and scale distributed systems that are already under pressure.

Improve performance, reliability, and AI workflows your team actually uses.

Book a consulting call →View selected work

20+

Years experience

25%

Throughput gain

50%

Faster delivery

What I do

Real production problems. Solved at the architecture level.

I work with startups and engineering teams facing real production pressure.

I work on systems that are already under pressure, not greenfield projects. Scaling bottlenecks, brittle architecture, weak engineering workflows, and slow delivery all trace back to decisions that need to change.

My work combines distributed systems architecture, technical judgment, and practical AI enablement so teams can make better decisions and move faster without losing reliability.

I work closely with teams to understand the real system, find what is slowing them down, and fix it.

Problems I help solve

If any of these are familiar, we should talk.

01

Your system is under real production pressure

The current architecture is starting to show cracks. What worked at earlier scale is now creating reliability risks, latency issues, and operational overhead.

02

Your team is shipping slower than it should

Technical direction, ownership, or workflows are not clear enough. Features take longer each quarter and the cause is not obvious from the surface.

03

You need to scale without creating new risks

Backend services, data pipelines, or event driven systems need to grow. The challenge is doing it without introducing new reliability problems or compounding existing debt.

04

You want to adopt AI across engineering

In a way that improves execution instead of adding noise. You need structure, standards, and measurable impact. Not demos and experiments.

Focus areas

Where I work. Technical problems, concrete outcomes.

01

Distributed Systems and Scale

Fix bottlenecks and improve system design for high throughput backend services, distributed systems, data pipelines, and multi tenant platforms under real production load.

  • Bottleneck diagnosis and resolution
  • System design and architecture guidance
  • Multi-tenant platform design
  • Event-driven and data pipeline architecture

02

Monolith to Microservices

Break down monolithic platforms into clearer service boundaries, safer communication patterns, and architectures that scale more cleanly over time.

  • Domain decomposition and boundary design
  • Communication pattern design
  • Migration roadmap and risk assessment
  • Independent deployment model

03

Architecture and Technical Direction

Clarify trade offs, reduce architectural drift, and define stronger technical direction with long term consequences in mind.

  • Architecture decision review and documentation
  • Trade off analysis
  • Technical direction for leadership and engineering
  • Architecture Decision Records

04

Engineering Effectiveness

Identify what is slowing delivery down: workflows, ownership, practices, decision making. Then fix it.

  • Delivery bottleneck diagnosis
  • Engineering workflow improvements
  • Development practice improvements
  • Clearer technical decision making

05

AI Enabled Development Workflows

Integrate AI into real development environments in a way that changes how teams actually build software: workflows, standards, commands, skills, internal tooling, and CI/CD automation around tools like Cursor, VS Code, and Claude Code.

  • AI workflow standards and adoption
  • Custom commands and reusable skills
  • Internal developer tooling
  • CI/CD automation integration

06

Deep Technical Reviews

Review systems, bottlenecks, architecture, and development processes to identify what is holding the team back and what needs to change.

  • Written architecture assessment
  • Bottleneck and risk identification
  • Prioritized improvement roadmap
  • Executive summary for leadership

Selected impact

Real systems. Measurable outcomes.

Situation

Cypago's platform was a monolith handling security compliance data from multiple third-party sources. Deployments were coupled, domain ownership was unclear, and the architecture could not support the pace of product growth. A strategic POC with Check Point was on the line.

What I did

  • Led full architectural decomposition, defining domain boundaries, data ownership models, and service communication patterns
  • Built a multi-tenant distributed data ingestion pipeline integrating multiple third-party security platforms
  • Implemented event-driven processing for reliable, lossless data flow across compliance workflows
  • Delivered a production-ready POC with Check Point within scope and timeline
  • Worked across multi-tenancy, secure integrations, and encryption sensitive workflows

Result

Monolith successfully decomposed into independently deployable microservices. POC with Check Point resulted in a commercial partnership. Teams could deploy and scale services independently for the first time.

Situation

Sisense's analytics platform was hitting hard limits on data pipeline throughput. Multi-tenant ingestion, transformation, and metadata management were bottlenecked in production. Development velocity across multiple teams had stalled due to a tangled data modeling layer.

What I did

  • Took full architectural ownership of multi-tenant data pipelines across ingestion and transformation layers
  • Redesigned data modeling and metadata architecture to decouple ownership across teams
  • Defined distributed processing patterns with strict tenant isolation
  • Deployed and operated containerized services on Kubernetes
  • Played a key role in hiring and multi-tenant platform design

Result

Pipeline throughput improved by 20 to 25% in production. Development cycle time reduced by 50% across multiple engineering teams. Architectural changes adopted broadly across the platform.

Situation

Joined during a difficult period of team attrition. Engineering capacity had dropped significantly and high volume systems needed continued development and stability to remain in production.

What I did

  • Created engineering teams from zero during a period of significant attrition
  • Led and managed development groups of 6 to 15 engineers
  • Supported and improved high volume systems processing around 10,000 events per second
  • Restored engineering stability and delivery capacity across the organization

Result

Engineering teams rebuilt from zero. Systems processing 10,000 events per second stabilized and continued development. Delivery capacity restored across the organization. Later acquired by McAfee.

Situation

Joined xWise as one of the company's earliest engineers. The platform needed to go from early stage to production scale, with reliable ad serving, engagement mechanisms, and real time event delivery for affiliate driven activity.

What I did

  • Improved ad serving capacity from 3,000 to 15,000 requests per second without stability issues
  • Built A/B testing capabilities for end users and decision mechanisms to improve engagement and merchant outcomes
  • Designed real time outbound event delivery for affiliate driven user activity at a time when webhook style integration patterns were still uncommon across product teams
  • Established core backend architecture patterns at an early stage of the company

Result

Ad serving capacity scaled 5x to 15,000 RPS. A/B testing and engagement tooling shipped to production. Real time event delivery enabled affiliate integrations that were uncommon in the product at the time. Later acquired by Playtech.

From colleagues

What people say who have worked with me.

“Avishay brought valuable input to system design discussions, often proposing practical, out-of-the-box solutions to complex technical challenges. He regularly shared new knowledge and insights with the team, which made working with him both effective and enjoyable.”

Yoav Feuerstein

Director of R&D · Cypago

Cypago

“Avishay is my go-to guy for any software engineering dilemma I encounter - whether it's a low-level design pattern or a system architecture. He can always provide a valuable input and inspiring insights. With great passion for clean code and keeping good practices.”

Ahimeir Cohen

Staff Software Engineer · Unity

Unity

“He constantly thinks of ways of improving the product and comes up with great ideas. Avishay never stops learning and improving - reading publications, reviewing other people's code, consulting with peers, taking courses. You get a great team member whom you can count on to be there at the time of need.”

Barak Mendelevich

Software Engineering Manager · Sisense

Sisense

AI enablement

AI Enabled Engineering

I help teams use AI to improve how they build software.

This includes defining workflows, setting standards, and integrating tools like Cursor, VS Code, and Claude Code into real development environments.

I also introduce automation around AI usage in CI/CD so teams can move faster with consistency and control.

AI becomes part of how the team actually works, not a side experiment.

01

Workflow standards

Defining how tools like Cursor, VS Code, and Claude Code fit into real development workflows, with clear standards rather than ad hoc usage.

02

Commands and skills

Creating reusable, team-level AI workflows that encode how your team writes, reviews, and ships code consistently.

03

Internal tooling

Building developer tools that fit your codebase, your stack, and the way your engineers actually work.

04

CI/CD automation

Integrating AI driven automation into the delivery pipeline to improve speed, consistency, and quality at scale.

How I work

Clear process. Defined scope.

01

Diagnose

Understand the system, the bottlenecks, the workflow, and the real sources of friction. I form opinions quickly and share them directly.

02

Guide

Define the architecture, the trade offs, and the technical direction clearly enough for teams to move with confidence.

03

Support

Stay close to the team to help the decisions hold up through execution, without turning the engagement into hands on staff augmentation.

I take on one or two clients at a time. The work gets full attention.

My focus is on better decisions and stronger architecture, not on replacing your engineering team.

Engagements are time-boxed by default. The scope can extend, but there is always a defined exit.

About

Avishay Zamir

Avishay Zamir

Distributed Systems & AI Engineering Advisor, Barcelona, Spain

I am a distributed systems advisor with over 20 years of experience building backend platforms, scaling data intensive systems, and improving how engineering teams work.

My work spans architecture, performance, team building, delivery processes, and practical AI adoption. I help teams make better technical decisions, improve execution, and build systems that hold up under real production pressure.

Based in

Barcelona, Spain (CET), Remote worldwide

Availability

Limited, 1 to 2 clients at a time

Logic test

98th percentile, Alva Labs

Book a consulting call→LinkedIn→

Let's talk

Let's talk.

If your team is dealing with scaling pressure, architecture complexity, or the challenge of adopting AI in a way that actually improves delivery, let's talk.

Book a consulting call→Send an email→LinkedIn→

I usually reply within 1 to 2 days.