Full Stack Engineer

Safeturf is an AI-driven commercial underwriting and 360-degree monitoring platform built for point-of-sale lending companies, commercial banks, and payment processors. The platform automates business due diligence (KYB, KYC, watchlist screening, adverse media, financial risk), accelerates merchant onboarding and activation, provides ongoing risk monitoring with real-time alerts, and generates customizable compliance reporting.

Under the hood, the system ingests business data from partners, runs AI-powered field mapping, transforms and enriches data across PostgreSQL and Snowflake, and includes a Claude-powered chatbot for querying business data via natural language.

The team is small, ships fast, and owns everything end-to-end — from infrastructure to product.

What You Will Do:

1. Build FastAPI endpoints for partner integrations and product features
2. Write and maintain Prefect flows: data import, AI mapping, monitoring, license management
3. Design PostgreSQL schemas, write Alembic migrations, optimize queries on Cloud SQL
4. Ship React + TypeScript features across the platform (import UI, monitoring dashboard)
5. Extend the Claude-powered chatbot: add MCP servers, tune prompts, improve session handling
6. Manage Docker Compose services, GKE deployments, and GCP infrastructure
7. Work with Snowflake for monitoring and warehouse queries
8. Maintain Keycloak auth, Twilio call scheduling, and partner integrations

Must-Have Requirements:

1. Russian - speaking
2. Python 3.10+ (3+ years): Production FastAPI experience. Comfortable with Pydantic models, async endpoints, and modular backend architecture. Has used Alembic or similar migration tools.
3. PostgreSQL: Strong SQL. Can design schemas, write complex queries, reason about performance, and manage migrations. Cloud SQL or managed Postgres experience is a plus.
4. React + TypeScript: Can build and ship frontend features independently. Familiar with hooks, TanStack Query (or React Query), component composition, and Tailwind CSS.
5. Docker: Daily Docker Compose user. Can write Dockerfiles, debug container networking, manage multi-service setups.
6. GCP + Kubernetes: Hands-on with GKE or similar. Can deploy services, manage Cloud SQL, work with GCS, and debug pod issues.
7. API design: Has built RESTful APIs consumed by external partners. Understands auth patterns, error handling, and clean request/response models.
8. Workflow orchestration: Experience with Prefect, Airflow, Dagster, or similar. Can design multi-step data pipelines with error handling, retries, and observability.
9. Self-sufficiency: Can take a business problem and turn it into a deployed solution without hand-holding. Two other engineers, no PM, no dedicated DevOps.

Nice-to-Haves:

1. Prefect 3.x specifically (we use it heavily)
2. Snowflake experience (querying, schema design, monitoring)
3. LLM integration: OpenAI API, Claude API, pydantic-ai, prompt engineering
4. MCP (Model Context Protocol) — we run custom MCP servers for the chatbot
5. Keycloak administration and OIDC/OAuth2 flows
6. Twilio (Studio flows, programmable voice)
7. Terraform or Pulumi for infrastructure-as-code