Aivar is an AI-first technology partner where cutting-edge technology meets industry expertise to supercharge your projects.
Experience: 4–8 years | 2+ years building data-intensive pipelines
Technical Focus: Own accelerator's 6-n stage processing pipeline end-to-end: Ingestion to Output. Build the connectors, PII/PHI safety layer, compliance audit trail, and multi-tenant isolation. Pure platform engineering — async Python, event-driven architecture, enterprise-scale reliability.
-
Design the core pipeline orchestration framework with async Python, supporting parallel/sequential stage execution with retry and dead-letter queues.
-
Build the connector framework for 50+ enterprise data sources (APIs, databases, file systems, SaaS platforms).
-
Implement safety validation — PII/PHI detection, data integrity checks, format compliance, input sanitisation.
-
Build the reconstruction module — transforming extracted data into Markdown/vector representations and context maps.
-
Design event-driven architecture with message queuing, backpressure handling, and graceful degradation.
-
Implement immutable audit trail logging across all 6 stages — every decision, action, and transformation.
-
Build multi-tenant data isolation, ensuring strict customer separation at every pipeline stage.
-
Distributed Python systems — asyncio, aiohttp, Celery; not scripting.
-
Message queues/streaming — Kafka, SQS, RabbitMQ, or Redis Streams.
-
High-volume data pipelines — 100K+ documents/records per batch.
-
API design — REST/GraphQL, versioning, rate limiting, and auth.
-
AWS services — Lambda, SQS/SNS, MSK, EKS, Step Functions, ECS/EKS, S3, DynamoDB, RDS.
-
Multi-tenant SaaS architecture and customer data isolation patterns.
-
Production habits — comprehensive testing, CI/CD, monitoring, and incident response.
Python, FastAPI, Celery, asyncio, Kafka/SQS/Redis Streams, PostgreSQL, DynamoDB, Redis, Elasticsearch, Terraform/CDK, Prometheus/Grafana/OpenTelemetry, GitHub Actions, Docker, AWS (Lambda, ECS, Step Functions, S3)
