Aelina Das

About

Backend engineer with recent Barclays experience building a Python asyncio + Kafka risk reporting service for trading desks, including a major refactor from blocking batch processing to event-driven incremental pipelines to restore intraday/EOD performance. Also shipped an applied AI feature using OpenAI fine-tuning to classify risk-breach severity and generate trader/risk-manager summaries with robust retry/fallback handling, plus demonstrated strong database/query optimization (triggers, materialized views, partial indexes) in a risk-limits/breaches domain.

Experience

Education

Key Strengths

• Debugged and resolved production performance degradation by identifying blocking batch step and refactoring to Kafka-driven incremental processing
• Strong testing discipline (documenting edge cases; unit/integration tests; load testing; gradual rollout via feature flags)
• Applied LLMs to automate risk-breach triage (severity classification, natural-language summaries, automated email notifications with remediation suggestions)
• Designed robust LLM failure handling (JSON parsing safeguards, retries, fallbacks to manual investigation to avoid pipeline crashes)
• Database performance optimization using triggers, materialized views, and partial indexes to eliminate N+1 subqueries and reduce scans
• Clear maintainable Python service architecture (separation of core business logic vs adapters/entrypoints; dependency management via pyproject.toml/lockfile)
• Reliability-oriented workflow/agent design thinking (idempotency, durable state, structured tool responses, circuit breakers, escalation, comprehensive tool-call logging)