Hi, I'm

Rıdvan Zengin

Senior Data & ML Engineer

Istanbul, Turkey

About

Data & ML Engineer with 5+ years of experience building data-intensive platforms and Proofs of Concept across real-world, large-scale systems. Experienced in designing end-to-end time-series pipelines, Digital Twin architectures — including AgriTwin (an open-source agricultural digital twin), IoTOps (an AI-native IoT operations platform), and a telecom network Digital Twin for a Tier-1 mobile operator — and scalable data models for complex domains. Strong background in developing ML/DL solutions, backend services, and full-stack tools, including RESTful APIs and web interfaces. Hands-on experience with containerized systems and cloud deployments, with a focus on turning ambiguous requirements into working, production-ready solutions through collaboration and technical ownership.

Projects

EnvelOps — AI Conversation Orchestration

An open-source multi-tenant AI orchestration platform that routes customer messages through a configurable pipeline with intent detection, RAG, tool calling, lead scoring, and safety-gated reply generation. Demonstrates configurable AI behavior per business without hardcoded vertical logic, backed by LangGraph workflows and pgvector-powered knowledge retrieval.

FastAPIReactTypeScriptPostgreSQLpgvectorLangGraphRAGCeleryRedisDocker

IoTOps — AI-Native IoT Platform

Visually configure MQTT/HTTP/Kafka/AMQP telemetry collectors, automate event-driven workflows with real-time and scheduled rule evaluation, and build drag-resizable dashboards — no hand-written Telegraf configs. An AI Co-pilot (Claude or Gemini, real tool-calling over live telemetry) answers data questions and suggests new automations or dashboards as reviewable drafts — never auto-created.

FastAPIReactTypeScriptTimescaleDBMongoDBRedisCeleryDockerClaudeGemini

AgriTwin — Agricultural Digital Twin

An open-source agricultural Digital Twin for precision farming that integrates satellite imagery, climate, soil, and economic datasets into an interactive H3-based spatial model for crop suitability analysis, what-if scenario simulation, and yield & profitability forecasting. Serves 346,000+ H3 hexagonal cells across Konya Province with per-cell crop suitability scores for 8 crops, async scenario re-scoring via Celery workers, and an interactive map built with MapLibre GL JS.

PythonGeoPandasPostGISH3TimescaleDBFlaskMapLibre GL JSCeleryRedisDocker

Skills & Tools

Languages
PythonSQLGolangFluxGit
Data & ML
PyTorchDeep LearningMachine LearningMLflowData Science
Databases
PostgreSQLpgvectorClickHouseInfluxDBTimescaleDBMySQLMongoDBCassandraRedis
Geospatial
PostGISGeoPandasH3MapLibre GL JS
Applied AI
LLM ApplicationsAI AgentsRAGLangGraphTool CallingPrompt EngineeringOllamaOpenAI-compatible APIs
Infrastructure
DockerKafkaSparkAWSLinuxRabbitMQCeleryGitLab
Tools
MQTTTelegrafGrafanaFlaskFastAPISNMPNetflow

Experience

Senior Data Engineer

Oct 2024 – Present

BTS Group

  • Led end-to-end development of a Digital Twin PoC for a Tier-1 mobile network operator — 4 use cases (Inventory Twin, Backhaul Topology Twin, Earthquake Impact Simulation, AI-based What-If simulations) across ~6,000 sites and 90,000+ network cells.
  • Primary technical contact for the customer, translating evolving business requirements into deliverable solutions under PoC constraints.
  • Architected a modular observability platform for real-time data collection and event detection across SNMP, HTTP, ICMP, MQTT, and Netflow.
  • Built custom Telegraf plugins (MQTT processor, JSON input, Redis enricher, Celery output) and a PoC system for on-demand Telegraf service deployment.
  • Designed a full-stack web platform for KPI definition, rule-based event detection, and custom dashboard generation for non-technical users.

Data & ML Engineer

Aug 2021 – Sep 2024

CRKTech

  • Developed AI-powered monitoring solutions for dairy farms by processing continuous IoT sensor data from livestock to improve animal health, welfare, and farm productivity.
  • Built end-to-end time-series data pipelines, transforming raw MQTT telemetry through parsing, feature engineering, and temporal aggregation for machine learning and real-time analytics.
  • Designed, trained, and deployed 5+ machine learning and deep learning models for production use cases including estrus detection and calving prediction.
  • Developed event-driven backend services and REST APIs using Flask to deliver real-time alerts and mobile notifications to farmers and veterinarians.
  • Built customer-facing monitoring applications integrating AI predictions, historical analytics, and operational insights for livestock management.
  • Managed Dockerized production services on AWS, ensuring reliable deployment, monitoring, and continuous updates.