Personal information hidden

Maksym Syhlovyi
[open contact info](look above in the "contact info" section) • [open contact info](look above in the "contact info" section) • Telegram • LinkedIn • GitHub

Python / AI Engineer with 3 years of experience building production LLM systems — LangGraph agents,
evaluation frameworks, Text-to-SQL, and cloud infrastructure on AWS

Experience Skills
ActlysAI FEB 2025 – PRESENT
Python
Built a LangGraph reminder-email agent and financial file classification / summarization
FastAPI
agent for a tax/accounting SaaS. Team: 5 BE, 2 FE, 1 BA, 1 PM
LangGraph
• Cut LLM costs ~80% by replacing LLM email generation with a template engine
LangChain
• Built an LLM column-classification step with Redis caching and an eval harness Flask
integrated in CI with Github Actions Django/DRF
• Architected agent-to-agent integration, integrated the reminder-email agent PostgreSQL
with the platform's main orchestrator SQL
• Integrate AWS Converse API to extract structured financial data from client files SQLAlchemy
Pydantic
Built a platform generating and publishing complete SEO sites end-to-end — LLM-written AWS
content, AI imagery, static rendering, automated GitHub + Cloudflare deployment — in Docker
single and bulk modes. Team: 2 BE, 1 SEO, 1 PM Redis
• Designed provider-agnostic LLM generation (OpenAI, Anthropic, DeepSeek, Celery
Gemini), allowing the best model to be picked based on SEO metrics TypeScript
• Freed up ~60% of QA review time with an audit service crawling deployed sites HTML/CSS
and producing pass/fail reports Linux
• Implemented a FastAPI CMS service: content updates, keyword link injection, CI/CD
redirects, subdomain cloning, domain substitution, and cloaking
• Built a meta title/description evaluator to drive prompt iteration
Languages
Stack: Python, FastAPI, LangChain, LangGraph, Celery, Redis, PostgreSQL, SQLAlchemy, Ukrainian — Native
Pydantic, AWS (AgentCore, Lambda, SQS, SNS, DynamoDB), Prometheus, Docker
English — B2
SoftwarePath JULY 2023 – JAN 2025 Slovak — A2
Built a Text-to-SQL LangGraph agent for querying users' data, extracted into PostgreSQL
by data pipelines. Team: 2 BE, 1 DS, 1 PM
• Built a Text-to-SQL evaluation dataset and evaluator to improve the prompt,
wired into CI/CD
• Extended the agent with chart-generation tools visualising query results in the
chat
• Integrated Stripe billing, Trello webhooks, and Google Drive API with OAuth2
authentication
• Developed data processing pipelines using Celery/Redis for data extraction

Stack: Python, FastAPI, Django, LangGraph, LangChain, TypeScript, JavaScript, Docker,
PostgreSQL, Redis, AWS (Lambda, SQS, ECS, RDS, EC2, S3), Azure OpenAI, Celery, Aiohttp

Education SEPTEMBER 2020 – MAY 2024

Bachelor of Transport Technologies | TNTY | Ternopil