Full-Stack Web + AI Engineer (LLM-agents)

Remote part-time position for a full-stack and AI engineer to develop an LLM-agent platform for messaging automation. Salary 60,000 RUB on hand, 4 hours/day.

We are developing an LLM-agent platform that, on behalf of the user, maintains correspondence in messengers. We solve nontrivial tasks at the intersection of AI, automation, and security. Tasks within the part-time day: - Backend and API for orchestrating LLM agents (dialogue memory, multilingualism, user profiles, escalation to operator); - Integration of agents with messengers: processing qualification, objections, transfer to a live operator; - Development of a client dashboard: onboarding, dashboard with AI quality metrics, user portrait; - Work with a device farm: automation on Playwright/Android, bypassing behavioral anti-detection (including for data collection and testing); - Ensuring security: data encryption, working with KMS, keeping access logs to sensitive information.

Expected: - Experience with Python in production from 3 years - confident mastery and architectural decisions; - Real commercial experience with LLM applications (from prototype to production); - React/Next.js at a level sufficient for independent development of the dashboard; - Practical experience with browser automation (Playwright/Puppeteer) and/or Android farm; - Understanding of antifraud systems and ways to bypass them; - Willingness to work with sensitive data, knowledge of encryption and key management approaches. Technical stack (mandatory minimum, but your suggestions are welcome): Python, FastAPI, PostgreSQL, Redis, LangChain, LangGraph, open-weight LLMs, pgvector/Chroma, React/Next.js, TypeScript, Playwright, Docker.

Conditions: - Fully remote format; - Flexible schedule - 4 hours per day (20 hours per week), can be split at your discretion within the working day (in agreement with the team); - Salary - 60,000 RUB on hand; - Level - Lead/Senior (high independence and responsibility); - Participation in the development of an internal AI ecosystem for a studio with 10+ active projects; - Project details and specific tasks are discussed at the interview.