Petros is available for hire

Petros Kalos

Verified Expert in Engineering

Back-end Developer

Location

London, United Kingdom

Toptal Member Since

September 12, 2022

Petros is a lead engineer with 9+ years of experience building high-performance, reliable, secure, and scalable systems. He is keen on back-end development, particularly machine learning, but can also deliver solid full-stack solutions. Petros has worked for different size companies in the multimedia (data compression), fintech (investment banking), and retail industries, delivering reliable and scalable code.

Back-end Linux Python Amazon Web Services (AWS)Software Engineering Distributed Systems Multithreading Amazon RDS Python 3 C++SQL Amazon DynamoDB Amazon S3 (AWS S3)Java TypeScript OpenCL MIPS

Portfolio

Amazon UK

Java, Amazon Web Services (AWS), Machine Learning, Python 3, Jenkins, Linux...

Morgan Stanley

C++, Python 3, Linux, Multithreading, Distributed Systems, TCP/IP

V-Nova Ltd

C++, Python 3, NumPy, SciPy, NVIDIA CUDA, SIMD, OpenCL, CodeXL, Oclgrind...

Experience

Multithreading - 10 years Distributed Systems - 10 years Linux - 10 years Software Engineering - 10 years Python 3 - 10 years C++ - 7 years Amazon Web Services (AWS) - 5 years Java - 3 years

Availability

Part-time

Preferred Environment

C++, Java, SQL, Amazon Web Services (AWS), Linux, NVIDIA CUDA, PyTorch, Python, Amazon SageMaker

The most amazing...

...project I've designed, developed, and productionized is a RoBERTa-based model currently serving millions of daily requests.

Work Experience

Senior Software Engineer and Tech Team Lead

2019 - PRESENT

Amazon UK

Designed and developed real-time and batch ML services (AWS and CDK) hosting models specifically trained to answer product questions using Amazon's vast product catalog information, reviews, and QAs like "Alexa, find the best headphones on Amazon."
Developed the full stack in a distributed infrastructure providing multi-region, low-latency, scalable, highly available, and data-sensitive solutions for new Alexa Shopping experiences using Java and Python.
Led a team that delivers customer experiences that help customers discover and learn more about the products listed on Amazon. Spearheaded large business-critical initiatives from design to delivery, applying a long-term architecture strategy.

Technologies: Java, Amazon Web Services (AWS), Machine Learning, Python 3, Jenkins, Linux, Back-end, APIs, Amazon RDS, Amazon DynamoDB, Amazon S3 (AWS S3), TypeScript

Technology Associate

2017 - 2019

Morgan Stanley

Built core services and trading engines for high-frequency, low-latency algorithmic trading, focused on data aggregation (Sybase and real-time market data) and unified APIs' distribution (C++, protobufs, and FlatBuffers) over a network/shared memory.
Involved in Morgan Stanley's OSS technologies through github.com/Morgan-Stanley in two projects: Treadmill (container/cloud), troubleshooting system-level issues, and Testplan (framework), frequently contributing to code and community engagement.
Collected new requirements from business units, production, and QA support while engaging with customers daily and during multi-region "Ready For Business" calls for critical changes (MiFID2).
Managed a multi-region, business-critical production infrastructure, including deployment, upgrades, decommissions, and security auditing.

Technologies: C++, Python 3, Linux, Multithreading, Distributed Systems, TCP/IP

Lead Software Engineer

2015 - 2017

V-Nova Ltd

Led the design and implementation of Perseus compression technology. Designed a high-performance multi-platform CPU and GPU solution using C++11, SIMD, CUDA (Jetson TK/X1 and Tesla), OpenCL1.1, GPUOpen CodeXL, Oclgrind, and NVIDIA Nsight.
Led the research for the next Perseus generations with NumPy, SciPy, and C++. Designed a highly efficient and highly parallel (GPUs) arithmetic coder with low memory bandwidth requirements.
Introduced continuous integration, coding standards, and documentation. Used Jenkins CI and integrated it with GitLab to automatically report coding standards violations (PEP 8 and ClangFormat), building and testing, and documentation generation.

Technologies: C++, Python 3, NumPy, SciPy, NVIDIA CUDA, SIMD, OpenCL, CodeXL, Oclgrind, NVIDIA NSight, Video Processing, TCP/IP

Software Design Engineer

2013 - 2015

Imagination Technologies

Performed C and C++ software development using video codecs and Linux Kernel.
Maintained and extended a heterogeneous distributed testing framework used to detect video coding failures across devices.
Developed video tests for Android using Java (MediaCodec, MediaPlayer, and ExoPlayer) and native (C++, Stagefright, OpenMAX, and Skia) APIs.
Built and deployed Android AOSP (x86, MIPS, and ARM), drivers' integration, and Kernel configuration.
Supported continuous integration using Jenkins, namely jobs management, script building, and automated testing.

Technologies: C++, C, Video Codecs, Linux, Android, Jenkins, MIPS, ARM

Experience

Machine Learning Model Productionization

Collaborating with a science team, I designed, developed, and productionized a RoBERTa-based model by optimizing it to fit strict latency budgets. The model was built using cloud-native AWS technologies like SageMaker for scaling and reliability, utilizing full CI/CD and precision validation. Currently, it is serving millions of requests per day.

gRPC Migration

I owned the migration of a high-throughput, low-latency trading-related service from a proprietary serialization/RPC protocol to Protocol Buffers (protobufs), FlatBuffers, and gRPC. The new approach made the system faster, more reliable, and easier to maintain and extend.

Massively Parallel Video Codec

I contributed to the science involved in entropy compression technologies and worked on the algorithms for developing a highly-efficient novel, one-frame latency video codec.

I built a scientific prototype for quick experimentation in Python based on SciPy, NumPy, and sklearn. I also developed the production-ready codec that utilizes off-the-shelf GPUs to achieve 4x4K encodings and decodings on a single machine.

Education

2008 - 2013

Master's Degree in Software Engineering

University of Thessaly - Volos, Greece

Skills

Languages

Python, C++, Java, Python 3, SQL, TypeScript, C, MIPS

Platforms

Amazon Web Services (AWS), Linux, NVIDIA CUDA, Android

Other

Back-end, Software Engineering, Distributed Systems, Multithreading, H264, APIs, Video Processing, Amazon RDS, Machine Learning, SIMD, Oclgrind, NVIDIA NSight, Video Codecs, ARM, Deep Learning, Transformers, RPC, H.265, TCP/IP

Storage

Amazon DynamoDB, Amazon S3 (AWS S3)

Frameworks

OpenCL, gRPC

Libraries/APIs

NumPy, SciPy, FlatBuffers, Protobuf, PyTorch

Tools

Jenkins, CodeXL, Amazon SageMaker

Paradigms

GPGPU

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring