Data Engineer · Cloud Architect · AI Builder · Creator of mygol.ai
Innovative Data Engineer with 5+ years building cloud‑native data platforms and privacy‑first APIs on GCP/AWS/Snowflake. Led cost and latency wins including $100K/month BigQuery savings and 87% delivery‑SLA reduction. I ship observable, scalable systems and pragmatic LLM/RAG features with guardrails.
- Cloud‑native data platforms on GCP (Cloud Run, Dataflow, Pub/Sub, BigQuery, dbt, Airflow, Spark)
- Privacy‑first APIs (GDPR/CCPA), observability with OpenTelemetry, and async event pipelines
- AI features: LLM/RAG, evals and hallucination mitigation for production reliability
- Now building: mygol.ai — an automated job‑apply engine on Cloud Run + Next.js
- Kafka Stock Streaming (TSLA prototype) — Stream stock price ticks from Polygon.io into Kafka, persist to Postgres, and compute SMA-based alerts. github.com/Ps-budd/kafka-stock-streaming
- Kafka Stock Streaming (TSLA prototype) — Stream stock price ticks from Polygon.io into Kafka, persist to Postgres, and compute SMA-based alerts.
github.com/Ps-budd/kafka-stock-streaming
Skills:
-
First‑Party Segment API (GCP) — Cloud Run + API Gateway + Pub/Sub/Dataflow; OpenTelemetry tracing; opt‑out & expiration endpoints.
Stack:GCP·Cloud Run·API Gateway·BigQuery·OpenTelemetry -
CRM Delivery Platform — PII encryption/salting; SFTP/GCS/S3 chunked delivery with retries; 2 days → 6 hours.
-
LLM Audience Segmenter — NL → audience segments; 90% manual effort ↓; 85% launch time ↓; guardrails + evals.
Stack:FastAPI·OpenAI/Gemini·Guardrails·Evals -
BigQuery Cost Optimizer — stored procedures + usage analyzer; partition/prune; ~$100K/month saved.
- Email: adi.dubey552@gmail.com
- LinkedIn: adityadubey09
- Resume: PDF
Always happy to chat about data platforms, cloud, and productionizing AI.

