20%
Platform cost reduction through optimization and governance work.
Berlin · Principal Data Engineer · AI + Platform Systems
Building reliable data and AI systems that scale in production.
I’m Kushagara, a principal-level engineer focused on distributed data infrastructure, platform engineering, cloud architecture, and agentic AI systems. I work across data platforms, developer tooling, system design, and production-grade delivery.
Selected impact
Production-scale systems, platform migrations, and engineering outcomes drawn directly from your career history.
500M+
Daily clickstream events handled with sub-second latency at BookMyShow.
50M+
Users supported through audience segmentation systems.
4+
Core domains spanning data, AI systems, cloud infrastructure, and platform engineering.
Work
Representative problems and systems that reflect the kind of engineering work you do today.
Governance and cost accountability
Drove a data platform FinOps initiative by aligning usage attribution, budgeting, and ownership for clearer accountability and more efficient utilization.
Agent reliability workflows
Contributed to an agent reliability initiative centered on structured evaluation, observability, and improving confidence in agent workflows.
CDC and ingestion modernization
Helped evolve ingestion from batch to near real-time CDC-based pipelines while improving maintainability across the wider platform.
Athena to Snowflake migration
Led the end-to-end migration of the warehouse stack from Athena-based architecture to Snowflake for a stronger analytical foundation.
High-volume event systems
Built and maintained streaming and clickstream systems at scale, including workloads above 500 million events per day with low-latency delivery.
Experience
A career spanning cloud platforms, distributed data systems, internal tooling, and product-facing engineering.
Solaris SE — Principal Data Engineer, Data Platform and Infrastructure
Leading data integrations, platform standardization, infrastructure modernization, and self-service capabilities across the organization.
- Led company-wide data and database standardization, data contracts exploration, and REST-based data exposure for internal teams.
- Scaled ingestion on Amazon S3, supported ML infrastructure on SageMaker, and improved orchestration with Airflow and ECS Fargate.
- Established dbt practices, CI/CD integration, documentation standards, and Elementary adoption.
- Built monitoring, alerting, and SLA tracking with CloudWatch and Grafana while supporting audits and anomaly detection systems.
Joonko AG — Data Engineer
First data engineer in the company, building the platform foundations from scratch.
- Built AWS-based data infrastructure and Kubernetes-native workflows using Argo.
- Helped introduce Kafka and built an API gateway for sending data into platform event streams.
- Designed reprocessing architecture for failed events and implemented lag monitoring, alerting, and engineering standards.
BookMyShow — Senior Software Engineer, Data
Worked across high-scale event systems, analytics infrastructure, and product engineering.
- Set up and maintained big data infrastructure including Hadoop, Kafka, Kafka Connect, Kafka Streams, Cloudera, and Spark workloads.
- Built a clickstream system handling more than 500 million events per day with sub-second latency.
- Contributed to dynamic pricing, audience segmentation, and earlier mobile web work including the BookMyShow PWA.
Hansa Cequity — Software Developer
Built campaign software, analytics workflows, and internal reporting tools.
- Developed web applications for marketing campaigns in collaboration with digital marketing and engineering teams.
- Built automated reporting systems using Python and contributed to internal Redshift and Redis-based products.
Stack
Practical technologies used across architecture, implementation, operations, and AI-oriented workflows.
Languages and APIs
PythonJavaGolangRustC++JavaScriptSQLFastAPIFlaskExpressREST APIs
Data and AI systems
SnowflakePostgreSQLMySQLCassandraAmazon S3dbtAirflowKafkaSparkFlinkCDC pipelinesRAGEmbeddingsSemantic retrievalTool callingAgent evaluation
Infra and operations
AWSDockerKubernetesTerraformJenkinsBashECS FargateSageMakerGrafanaCloudWatchFinOpsGovernanceSystem design
About
I’m based in Berlin and work across data engineering, platform architecture, cloud infrastructure, and AI-enabled developer systems. My background spans early platform building, large-scale event systems, warehouse migration, infrastructure automation, and practical application of modern AI workflows.
Notes on systems, AI agents, and engineering practice.
Thoughts, Ideas, Architecture breakdowns, Migration lessons, Experiences, Learnings and Experiments
What changed when we moved from batch ingestion to CDC
A practical write-up on reliability, schema evolution, backfills, consumer contracts, and where the operational complexity really moved.
What makes agent evaluation actually useful in production
Lessons from turning agent behavior into measurable workflows using structured traces, goals, plans, actions, and failure analysis.
FinOps for data teams beyond monthly cost dashboards
How ownership, attribution, and engineering accountability can reshape platform decisions over time.