Hayk Harutyunyan, Developer in Yerevan, Armenia
Hayk is available for hire
Hire Hayk

Hayk Harutyunyan

Bio

Hayk is an engineering leader combining infrastructure expertise (Kubernetes, AWS) with applied AI engineering. He is experienced in productionizing agentic systems: moving beyond demos to build reliable, observable, and cost-efficient AI architectures. Hayk has a unique background in FinOp and cost-aware system design.

Portfolio

Relaunch
Large Language Models (LLMs), Artificial Intelligence (AI), React, Python...
nOps
Python, Django, PostgreSQL, Microservices, Kubernetes, Apache Kafka, Docker...
BMAT
Python, Django, PostgreSQL, APIs, REST, Linux, Bash, Celery, Redis Queue...

Experience

  • Python - 8 years
  • React - 5 years
  • Amazon Web Services (AWS) - 5 years
  • Django - 5 years
  • Microservices - 4 years
  • APIs - 4 years
  • FastAPI - 4 years
  • SQLAlchemy - 2 years

Preferred Environment

Vim Text Editor, Amazon Web Services (AWS), Claude Code

The most amazing...

...thing I built was a user behavior simulation engine for A/B tests that achieved over 80% correctness when compared with real world results.

Work Experience

Co-founder

2025 - PRESENT
Relaunch
  • Built a production agentic AI orchestration system at Relaunch with planner/executor/evaluator workflows, tool use, checkpointed resumption, multi-model routing, and Postgres-backed state.
  • Architected a large language model (LLM) user behavior prediction engine, successfully modeling statistical outcomes of A/B tests with over 80% correlation to real-world data.
  • Built a robust evaluation framework to benchmark simulated user fidelity against real-world analytics data.
  • Developed a low-latency retrieval-augmented generation (RAG) subsystem for the product chat interface, optimizing retrieval precision on proprietary experiment datasets.
Technologies: Large Language Models (LLMs), Artificial Intelligence (AI), React, Python, AI Agents, Agentic AI, LangGraph, Agentic AI Systems, Software Architecture, Amazon Web Services (AWS), Google Cloud Platform (GCP), Claude API, Gemini API, Cloud, CTO, RAG Systems, Retrieval-augmented Generation (RAG), Model Context Protocol (MCP), Infrastructure as Code (IaC), Claude Code, Code Review

Senior Software Engineer

2022 - 2024
nOps
  • Led a team of six to create a Kubernetes workload optimization system that chooses the most cost-optimal instances for a given set of workloads, consisting of a tool control dashboard together with a complex back end of APIs and automations.
  • Built several microservices from scratch and contributed to others, leveraging FastAPI, Kafka, and Kubernetes for inter-service communication and scaling.
  • Contributed to designing and implementing distributed data processing pipelines, such as event-based near real-time (NRT) ingestion and processing of AWS or Kubernetes metadata.
Technologies: Python, Django, PostgreSQL, Microservices, Kubernetes, Apache Kafka, Docker, CI/CD Pipelines, Back-end, Amazon Web Services (AWS), SQL, Data Engineering, GraphQL, REST APIs, FastAPI, SQLAlchemy, Full-stack, React, Software Architecture, Leadership, Infrastructure as Code (IaC)

Software Engineer

2020 - 2022
BMAT
  • Contributed to developing a complex data processing pipeline leveraging the capabilities of AWS Batch and Python's multi-core processing and threading libraries.
  • Built the pipeline's data reporting foundations using domain-driven design (DDD) principles.
  • Contributed to operating and maintaining said data processing pipeline, launching and monitoring queues, troubleshooting issues, and fixing bugs.
Technologies: Python, Django, PostgreSQL, APIs, REST, Linux, Bash, Celery, Redis Queue, MongoDB, Redis, CI/CD Pipelines, DevOps, Amazon Web Services (AWS), API Design, Distributed Systems, Cloud Architecture, SQL, Data Engineering, REST APIs, JavaScript, React

Experience

A Data Processing Platform

http://pronto.bmat.com
A Django-based platform that parses, ingests, and processes large volumes of file-based data with multiple processing steps and varying configurations per customer.

I was one of the back-end engineers maintaining and improving the pipeline. I leveraged the compute capacity of AWS Batch and Python multiprocessing libraries to build a performant system.

A Cloud Optimization Tool

http://nops.io
This project consisted of a distributed microservices-based SaaS application for cloud cost optimization. I was involved as a senior software engineer and built from scratch and contributed to significant platform features, such as data ingestion and processing pipelines, RI management automation tools, and visualization dashboards.

Relaunch

http://relaunch.ai
Relaunch.ai is an AI-powered growth platform designed to help businesses—especially product teams—analyze and optimize how their products grow and perform. It acts as a “co-pilot” for product growth, providing data-driven insights and simulations to improve key product metrics.

Education

2017 - 2018

Bachelor's Degree in Liberal Arts

University of Oxford - Oxford, UK

2014 - 2018

Bachelor's Degree in Liberal Arts

Middlebury College - Middlebury, Vermont, USA

Certifications

FEBRUARY 2023 - FEBRUARY 2025

FinOps Certified Practitioner

FinOps Foundation

JANUARY 2023 - JANUARY 2026

AWS Certified Developer

Amazon Web Services

Skills

Libraries/APIs

Django ORM, Pydantic, Google Ads API, SQLAlchemy, Rasa NLU, Redis Queue, Intercom API, REST APIs, Pandas, OpenAI API, React, Claude API

Tools

Docker Compose, Celery, Rasa.ai, Plotly, Pytest, Claude Code, Claude, Codex

Languages

Python, SQL, Bash, Python 3, GraphQL, Go, Swift, JavaScript

Frameworks

Django, Django REST Framework, Flask, LangGraph, OAuth 2

Paradigms

REST, Microservices, Continuous Delivery (CD), Continuous Integration (CI), DevOps, RESTful Development, Model Context Protocol (MCP)

Platforms

Amazon Web Services (AWS), Linux, Docker, Kubernetes, Apache Kafka, Google Cloud Platform (GCP)

Storage

Relational Databases, PostgreSQL, Redis, MongoDB, Elasticsearch

Other

APIs, Back-end, Distributed Systems, Cloud Architecture, API Design, Integration, Facebook Ads, FastAPI, CI/CD Pipelines, SDKs, Natural Language Processing (NLP), Chatbots, Data Visualization, Data Engineering, Artificial Intelligence (AI), Cost Reduction & Optimization (Cost-down), FinOps, Generative Pre-trained Transformers (GPT), Large Language Models (LLMs), Large Language Model Operations (LLMOps), Voice Chat, Full-stack, AI Agents, Agentic AI, Agentic AI Systems, Software Architecture, Gemini API, Leadership, API Integration, LangChain, Cloud, CTO, Engineering, RAG Systems, Retrieval-augmented Generation (RAG), Infrastructure as Code (IaC), Code Review

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.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

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