Igor Makhtes

Jan 2026 — Present Open Source

Creator & Lead Architect

Shep · Open Source shep.bot github.com/shep-ai/shep

Created and architected Shep, an innovative open-source AI Software Development Life Cycle (SDLC) platform designed to automate and orchestrate complex engineering workflows using autonomous agents.

Spec Driven Development: Pioneered an architecture where high-level architectural specifications are dynamically translated into executable code, automated tests, and infrastructure via context-aware AI agents.
Agent-Agnostic Execution Engine: Designed a highly modular core utilizing LangChain and LangGraph for graph execution, supporting plug-and-play integration of diverse LLMs (OpenAI, Anthropic, Gemini, local models) without vendor lock-in.
Local-First Execution: Engineered the platform so that all orchestration, data processing, and graph execution run fully locally, with external network calls restricted solely to LLM inference.
Parallel Orchestration: Engineered concurrent execution pipelines allowing multiple agents to tackle isolated development tasks simultaneously (e.g., writing tests while generating business logic), drastically reducing delivery times.
Self-Healing Workflows: Implemented iterative feedback loops where agents automatically compile code, run tests, analyze failure logs, and patch errors until the specification is successfully met.

2024 — Jan 2026 Staff Software Engineer 2023 — 2024 · Senior Software Developer

Staff Software Engineer

Senior Software Developer · 2023 — 2024

Snyk

Unified Assets Platform

Led the technical plan and implementation of the unified assets platform, combining legacy Snyk data processing and UI/UX into a single new architecture. Built the technical architecture for a large multi-team effort, initiating the hands-on implementation.

Data Engineering & Scale

Achieved a 500x platform scale-up by redesigning the ETL process workflow and implementing MongoDB sharding, enabling massive data processing capacity with only a 2-3x increase in infrastructure costs, ensuring cost-effective scaling.
Led the integration of Apache Flink stream processing, transitioning from batch ELT periodic pipelines to real-time data streaming. Designed and implemented the ingestion and processing layer architecture, allowing for easy deployment of new stream-processing applications. Facilitated team adoption by mentoring team members on the use of Flink and developing stream-processing solutions. Built comprehensive CI/CD pipelines, alongside monitoring and alerting systems, to ensure smooth and reliable operation of real-time data streams.
Engineered a robust ELT pipeline for data synchronization, extraction, processing, and integration, leveraging kafka-connect for scalable real-time data ingestion, Debezium CDC to track and capture database changes, and DBT (SQL) for modular transformations. Integrated with Snowflake to enable high-performance analytics and data warehousing.

Enso.Security Integration

Led the migration, integration and deployment of enso.security platform into Snyk ecosystem (Snyk AppRisk) including:

CI/CD migration and adaptation (CircleCI)
Projects deployment infrastructure (Helm charts & Terraform)
Implementing and adding missing functionalities (kafka-connectors, kafka brokers, CDC)
API-gateway (gloo), AuthN, AuthZ integration
DBs migrations
DataDog monitoring and alerting setup
SSO integration (Okta)

2022 — 2023

Tech Lead

Enso.Security (Acquired by Snyk)

Led the design and architecture of the system, overseeing infrastructure and technology integration.
Integrated and implemented new technologies such as ArgoCD, Argo Workflows, Nx Monorepo and NestJS Framework.
Refactored existing pipelines for data extraction and processing using Argo Workflows and in-cluster agent micro-service.
Optimized system components, significantly improving system reliability through enhanced e2e testing and refactored CI/CD pipelines.
Enhanced observability and monitoring with a full-stack integration of Prometheus for metrics collection, Jaeger for distributed tracing, and Grafana dashboards for real-time performance monitoring and alerting.
Introduced and applied Domain-Driven Design (DDD) principles in the system architecture to decouple services, fostering better scalability, maintainability, and modularity in a microservices environment.
Actively worked with MongoDB and ArangoDB for handling complex data storage needs, utilizing Fastify and NestJS frameworks in a TypeScript-based ecosystem.

2019 — 2022

System Architect

Cogniteam LTD · cognimbus.com

Architecture design, development and teams management of cognimbus.com. Hands-on experience:

Back-end: C#, ASP.NET Core 5.0/6.0, Kestrel (Linux), Entity Framework (.NET ORM), NEST (Elasticsearch client), TUS (.NET file server). REST/GRPC/Websocket, HTTP2/HTTP3. Serilog (Structured logging). JWT authorization. SendGrid/Twilio.
Databases: MySQL, Elasticsearch, MongoDB.
Front-end: Developed data streaming and message processing module. UI design and prototyping (Photoshop & Figma). React (TypeScript), ThreeJS (react-fiber), Redux, MaterialUI. Background data processing using webworker. Service worker.
Agent: C# cross-platform (Linux/Windows/MacOS) cloud connected application running as system service. RTI Connext DDS. LiteDB. CLI (including autocomplete) development. Bash scripting. Mediatr (mediator pattern implementation). Public SDK development. Unity3D (WebGL).
DevOps: Setup, management and monitoring of servers and CI/CD infrastructure. Oracle Cloud. CI/CD Pipelines: Gitlab CI, Github Actions, Unit/Integration tests pipelines, artifacts repositories (npm, nuget, docker, apt). Kafka, Redis, Vagrant, Docker, Docker-compose, Sentry, ELK, K8s, Helm, ArgoCD. Nginx (static pages server and reverse proxy load balancer). DNS & TLS certificates management, Certbot (LetsEncrypt).

2017 — 2019

Software Architect

Cogniteam LTD

Development management and hands-on tech lead for customer projects.

Linux Kernel, UBoot bootloader
Computer vision and deep learning using Tensorflow / Tensorflow Lite: Mobilenet object detection networks, Segmentation networks, Deep network training and optimization for low performance platform
CI/CD pipelines: Jenkins, Gitlab
Cross-platform CI/CD build for Linux/Windows, AMD64/ARMv7/ARMv8 architectures
AWS build and test pipeline (AWS RoboMaker, Lambda, CloudWatch, Cloud Formation)
Decision making system development
UI design & Web interface (Bootstrap / JQuery)

2016 — 2017

Senior Software Engineer

Cogniteam LTD

Developed different robots from the ground up including:

PCB prototyping (soldering and wiring)
Firmware development (Arduino), Wire communication protocol, Drivers
High-level management software
Web UI: Mapbox, ThreeJS, JQuery
Gazebo simulation

2016 — 2017

Senior Software Engineer

Bladeranger LTD · www.bladeranger.com

Pleco robot prototype design and development
C++, Python, ROS
3D Design and animation of promotional content (Blender)

2013 — 2016

Software Engineer

Cogniteam LTD

Developed algorithms for autonomous robots.

C++, Python, ROS/ROS2, Linux, BASH
Computer vision (OpenCV)
Mapping and localization algorithms
Navigation algorithms
Sensor data processing, filtering and calibration: 2D/3D lasers, IMU (gyro, accelerometers, magnetometers/3d compass), RGB/RGBD cameras, Sonars

2008 — 2010

Web Developer

Freelancer

Business web sites development
HTML / JS / JQuery / CSS

2005 — 2008

Software Developer

IDF

Software developer at military police headquarter
ASP.NET (C#)
C# / VB.NET Desktop application (WinForms)
Oracle DB, MS SQL DB

Creator & Lead Architect

Staff Software Engineer

Tech Lead

System Architect

Software Architect

Senior Software Engineer

Senior Software Engineer

Software Engineer

Web Developer

Software Developer

AI & Agent Stack i.

Core Competencies ii.

Languages i.

Education ii.

Bachelor of Science

Workshops & Talks iii.

Interests iv.