
Pavel Puzikov
Verified Expert in Engineering
Full-stack Developer
Lisbon, Portugal
Toptal member since December 16, 2014
Pavel is a full-stack engineer with 15+ years of experience building startups and enterprise software. He specializes in React, TypeScript, Node.js, Python, Ruby, APIs, cloud architecture, and AI workflows. Recently, he built AI document parsing workflows and logistics tools for autonomous delivery, and led refactors with Claude Code. As a founding architect at Zindi, a 70,000-member AI platform backed by Microsoft, Amazon, and Google, he also co-founded Flocktory, acquired for $20 million.
Portfolio
Experience
- JavaScript - 17 years
- Full-stack - 15 years
- REST APIs - 12 years
- Architecture - 12 years
- Node.js - 10 years
- React - 10 years
- TypeScript - 8 years
- Artificial Intelligence (AI) - 7 years
Preferred Environment
Vim Text Editor, Claude Code, Codex, GitHub
The most amazing...
...thing I've built is Zindi, Africa's largest AI competition platform.
Work Experience
Front-end Engineer
Gatik AI Inc.
- Worked on a platform for managing autonomous truck deliveries, supporting delivery planning, tracking, and coordination for internal operations.
- Owned and independently executed a ground-up refactor of a core internal application, improving performance and modernizing architecture without disrupting operational processes.
- Used Claude Code heavily as part of an AI-assisted engineering workflow to accelerate codebase analysis, refactoring, and implementation.
Software Architect
SmartChainServices
- Architected Azure AI-powered workflows to parse oil and gas contracts and transform unstructured document data into structured operational processes.
- Implemented a mathematical model to estimate remaining diesel volume from unreliable sensor readings in uneven containers.
- Designed and launched two revenue-driving products: workforce bonus calculation and diesel tracking.
Senior Software Developer
Distribusion Technologies GmbH
- Built and launched B2B transport operator integrations across Europe and South America.
- Refactored the data layer of a high-scale transport platform, reducing infrastructure costs by $200,000/year.
- Worked across APIs, PostgreSQL, MySQL, Docker, Kubernetes, Kafka, observability, and cloud infrastructure to improve production reliability.
Team Lead
Array, LLC
- Led architecture and delivery for a fintech product launch, managing six engineers across API design, integration, and secure data flows.
- Coordinated with external and internal engineering teams to align authentication, banking portal integration, and audit requirements.
- Supported external security audit readiness and implemented secure authentication/data-handling workflows.
Co-founding CTO
Zindi.africa
- Designed and implemented the MVP in six months, working with founders from the start.
- Built a React front end with custom data management and SSR to ensure time-to-interactive under 1.5 seconds.
- Engineered the API back end using Ruby on Rails and Postgres, with an average response time of 50ms, alongside a Python service for ML scoring.
- Managed cloud deployment on AWS and Azure with CI/CD, failover, and autoscaling, maintaining 99.99% uptime.
- Cultivated a community of 70,000 active data scientists and hosted competitions worth $500,000 in prizes.
- Led continuous development for five years, including a redesign in collaboration with an external agency.
Full-stack Developer
Tecstone Granite USA, Ltd.
- Built a browser-based 3D editor for creating granite monuments using WebGL and Three.js.
- Defined the requirements and architecture in a new and unexplored field.
- Developed the version that supports live editing, including changing materials and engravings, and importing from the company's catalog.
Full-stack Developer
Infraplan
- Developed an internal platform for an urban planning company that utilizes statistical analysis, built with Ruby on Rails and React.
- Inherited an outdated Ruby on Rails codebase, optimized it, and added a React front end.
- Implemented a data pipeline to support comprehensive data calculations and visualizations.
- Implemented data planning and visualization tool using Three.js.
Co-founder
Eleos
- Started a company to pursue the idea of an application that uses environmental signals to find the best time to deliver the message.
- Developed a web application, a mobile application, and a Facebook messaging bot as we experimented with the medium.
- Designed and built a back end using Ruby and Apache Kafka and trained an AI model for background signal processing.
Principal Architect
Votecastr (with Slate.com)
- Engineered a scalable back end to support a real-time election tracking and prediction engine, including a mobile app and website, within three months.
- Developed a real-time map widget for slate.com, handling over 20 million page views and 120,000 RPM.
- Coordinated with the data team and led the platform through election day/night in the Vice War Room.
Full-stack Developer
Predictions Dashboard
- Created an app to process and present vast amounts of data in a readable way. The data was hundreds of gigabytes of historical information with predictions for the future, delivered weekly in CSV format.
- Selected Google BigQuery for data storage, which provided linear query time, and created a pipeline to automatically update the data whenever new files were added to an Amazon S3 bucket.
- Built a Ruby on Rails back end with two-factor authentication for a data API and user management panel.
- Experimented with 2D and 3D data visualization techniques using Three.js, D3.js, and Cytoscape.js. Developed a dynamic front-end interface with React and D3.js.
CTO
Flocktory
- Joined as a CTO on the first day and built the MVP with Ruby on Rails in two months.
- Led development for two years and grew the engineering team to eight people.
- Contributed to achieving profitability and a user base of 5,2 million MAU within two years, culminating in a later $20 million acquisition by Nasdaq-listed QIWI.
Web Developer
Grapevine
- Founded and ran a web consultancy specializing in early-stage startups.
- Consulted with multiple clients, including habr.com with 13 million MAU and rutube.ru with 20 million MAU.
- Worked with the founders from inception, making prototypes, building the MVP, hiring, and transferring a project to an in-house team.
Experience
Zindi Africa
https://zindi.africaIntcitium
Votecastr
https://en.wikipedia.org/wiki/Votecastrezyorganizer
Education
Bachelor's Degree in Computer Science
Saint Petersburg State University of Information Technologies, Mechanics, and Optics - Saint Petersburg, Russia
Skills
Libraries/APIs
D3.js, Node.js, React, REST APIs, Google Maps API, Google Maps, Claude API, WebGL, Three.js, OpenStreetMap API, Apple Maps API, Bing Maps API, Mapbox API, Leaflet
Tools
Figma, Git, GIS, Terraform, GitLab CI/CD, Claude Code, Codex, GitHub, Grafana, Microsoft Power BI, Vim Text Editor, Sentry
Languages
CSS, TypeScript, HTML, Ruby, JavaScript, HTML5, ECMAScript (ES6), GraphQL, SQL, Python, Clojure, Elm, R
Frameworks
React Native, Redux, Ruby on Rails (RoR), Next.js, Tailwind CSS, NestJS, ClojureScript
Paradigms
Web Performance Optimization (WPO), DevOps, Functional Programming, REST, Agile
Platforms
Azure, Amazon Web Services (AWS), Heroku, Kubernetes, Azure AI Studio, iOS, Android, Google Cloud Platform (GCP), Docker, Jupyter Notebook, Mobile, Twilio, Blockchain, Apache Kafka, Firebase, Mapbox
Storage
MySQL, PostgreSQL, NoSQL, MongoDB, PostGIS, BigTable, Datadog
Other
Data Visualization, Software Engineering, Architecture, System Design, Full-stack, API Design, API Integration, Documentation, WebSockets, Back-end, APIs, Integration, Scraping, Technical Architecture, Front-end, Startups, CI/CD Pipelines, User Interface (UI), Geolocation, 3D, Terraform Cloud, AI Integration, AI Model Integration, AI Engineering, Agentic AI, AI-assisted Development, Systems Design, Artificial Intelligence (AI), Minimum Viable Product (MVP), Agentic AI Systems, AI Agents, AI Architecture, Venture Design, Progressive Web Applications (PWAs), Data Science, Machine Learning, Data Analysis, Storybook, Forecasting, GeoJSON, Maps, Prometheus, Mathematics, Document Parsing
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
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring