Real-Time AI Conversation Platform: POC → Production
Transformed a basic proof-of-concept into a secure, scalable, real-time AI
conversation-analysis product. Delivered to the first customer in six months;
scaled from 100 to 15,000 calls per day across multiple customers, integrations and regions.
High test coverage: unit, E2E, and scale tests throughout.
Python / async
WebSockets
SIPREC
LLM
AWS
ECS
ALB
VPC
S3
SQS
Engineering Excellence: From Team Building to Best Practices to Architecture
Worked in diverse roles as standalone developer, Team Leader and Architect, led work across diverse technologies: async microservices, RabbitMQ pipelines,
AWS Lambda + SQS infrastructure, C++ Build Systems. Established all important good practices, uniform work environments reducing friction,
code quality standards, and CI/CD infrastructure and design, including tests at all levels: E2E, Unit and Scale Tests.
Led routine development work as well as major architectural initiatives, including research, technology selection, and overseeing implementation by other engineers.
AWS
Microservices
Build systems
TDD
RabbitMQ
CI/CD
Architecture
WiFi Security Product: Core Engineering from the Ground Up
Engineered the core of a WiFi security product from scratch: from
low-level networking and WiFi monitoring to analysis tools,
Embedded Linux system configuration, and comprehensive testing
infrastructure. Handed off a production-ready prototype system
supporting a wide range of cyber security applications, with robust
tooling, orderly and efficient deployment capabilities to edge
devices, with security built in from the start.
C hacking
Python
802.11
Embedded Linux Distributions
TCP/IP
Networking
Low Level
Restoring Order in a Convoluted Build System
Took ownership of a large, convoluted CMake-based build system and restored order:
removed duplicated and overly complex logic, eliminated dependency on legacy Python versions,
and introduced build-specific tailored virtual environments, untagling misconfiguration
of Finder Modules, e.g. FindPython3, resolving missing Python.h headers etc.
Resolved deep-rooted sensitivity to dependence on legacy, stateful and opaque development servers.
Decomposed a convoluted both native and cross compiling scenario into a much cleaner, more modular and understandable structure.
Maintained constant communication with the customer's team and tech leads throughout to verify impact at every step.
CMake
C++
Python
Build Systems
Refactoring
SDLC
Technical Leadership at a Defensive Cyber Security Startup
Shaped both the product and the engineering culture at a defensive cyber security startup.
Designed the AWS Lambda + SQS backend infrastructure, defined company-wide
testing guidelines (E2E, integration, unit), and established a Standard Development Environment
that eliminated whole classes of environment-dependent bugs, while at the same time allowing developers to keep their favorite tools.
Negotiated with Product to translate product requirements into verifiable, fact-based tests. Led research and prototyping of new technologies, oversaw major technical transitions, and mentored the team on code quality standards and practices.
Research, prototyping and embedding of technologies including: S3, AWS Aurora, SQS, Lambda, Celery, docker, Gitlab-CI, networks
AWS Lambda
SQS
Python
Microservices
RabbitMQ
TDD
Architecture
Docker
GitLab CI