AI Data Engineer @ Meta | Building data pipelines & AI infrastructure at scale

• 🔧 Data Engineer at Meta Platforms, working on marketing analytics infrastructure and data clean rooms
• 🤖 Deeply interested in AI Engineering — I believe the future of data engineering is AI-native, where data is self-serve through natural language and pipelines build, monitor, and document themselves
• ☁️ Cloud-first mindset — I enjoy architecting on AWS and designing systems that are private, governed, and scalable by default
• 🌱 Currently exploring agentic AI workflows, LLM evaluation methods, and AI-assisted developer tooling
• 💬 Always happy to chat about data engineering, GenAI infrastructure, or making messy data trustworthy

• 🏅 AWS Certified Solutions Architect — Verify credential

🏗️ Data Pipeline Engineering — Production ETL/ELT pipelines processing high-volume datasets: 971M+ row identity refactors, multi-stream clean room ingestion, incremental backfills, and dimensional modeling with Spark, Airflow, Hive, Presto, and SQL — backed by data validation, SLA monitoring, and pipeline observability

🤖 AI Infrastructure for Data — Semantic models that let AI agents answer natural-language questions over warehouse data (drove LLM query accuracy from 66% to 90% via eval-driven development); Claude Code plugins for pipeline scaffolding, automated health monitoring, and AI-powered team onboarding

🧠 ML & LLM Engineering — Fine-tuning LLMs (BART, GPT-3.5) with PyTorch/CUDA, prompt engineering, RAG pipelines with FAISS embeddings, BERTopic modeling on multi-million record datasets, and ML-based anomaly alerting on production metrics

☁️ Cloud Data Architecture — Privacy-safe data infrastructure on AWS (S3, Glue, KMS, Kinesis, SageMaker) with encryption, pseudonymization, data governance, and cross-platform identity resolution; streaming and batch architectures on GCP (BigQuery, Dataflow, Pub/Sub)

📈 Metrics & Observability — Centralized metrics governance (51 metrics standardized across 6 domains), automated DoD/WoW/MoM anomaly detection, and self-serve analytics dashboards for cross-functional stakeholders

💬 Always open to conversations about data engineering, AI infrastructure, and building systems that make data self-serve. Reach out on LinkedIn or email me at srenikunta19@gmail.com!