Open to opportunities

Sonya Ling

@sonyaling

Message

Staff software engineer building scalable agentic AI and RAG systems for enterprise workflows.

United States

Message

What I'm looking for

I’m looking for a role where I can build scalable agentic AI and RAG platforms, own architecture and deployment, and drive observability/evaluation—partnering with teams to support human-in-the-loop workflows.

I’m a Staff Software Engineer with 13+ years leading architecture and design of scalable, high-throughput distributed systems and AI-powered platforms. I specialize in agentic orchestration frameworks and retrieval-augmented generation (RAG) systems that integrate with enterprise workflows, including memory architectures and human-in-the-loop collaboration.

I’ve built and shipped multi-agent solutions end-to-end—from rapid prototypes to production deployments. Recently, I designed and deployed a scalable AI-powered Prelegal SaaS using Claude Code, FastAPI (async), and Terraform, and delivered a full-stack solution under two weeks. I also developed a multi-agent orchestration system featuring planner/reporter/charter/retirement agents with AWS Lambda hosting, SQS queues, Aurora Postgres, and observability via OpenTelemetry and LangFuse.

Built an Agentic RAG system for healthcare domain using embeddings, ChromaDB vector store, hybrid semantic and keyword search retriever, reranker with Cross-Encoder, and LangGraph workflow orchestration. Incorporated routing agent, relevancy check agent, fallback web search, streaming structured output, tool callings, and citation requirements to enhance quality and prevent hallucination.

I bring strong leadership and engineering rigor to fast-paced teams. At OpenX, I led a Lookalike (LAL) data pipeline on GKE with Vertex AI integrations, optimizing inference jobs to reduce runtime by 70% and tuning BigQuery feature store queries to eliminate failures. I’m equally comfortable strengthening distributed systems—Kafka and streaming pipelines, cache concurrency and decryption in Go, and performance tuning—while ensuring reliability, evaluation, and traceability for AI systems.

Experience

Work history, roles, and key accomplishments

Current

Hands-on AI Projects

Current

N/A

Dec 2025 - Present (6 months)

Designed and deployed a scalable Prelegal SaaS using Claude Code, async FastAPI, and Terraform, integrating JWT auth (Clerk) and migrating storage to PostgreSQL; delivered a full-stack solution in under two weeks. Built agentic orchestration and healthcare RAG workflows with multi-agent routing, streaming structured outputs, tool calling, and OpenTelemetry/LangFuse observability.

Claude Code fastAPI Terraform Google Cloud Run Multi agent orchestration Agentic RAG LangGraph OpenTelemetry Langfuse PostgreSQL

Staff Software Engineer

OpenX

Mar 2023 - Oct 2025 (2 years 7 months)

Led architecture and development of the Lookalike (LAL) data pipeline, orchestrating TypeScript/GraphQL services on GKE with Pub/Sub and Vertex AI integration. Optimized LAL inference jobs to cut runtime by 70%, tuned BigQuery feature store queries for reliability, and monitored low-latency systems with Prometheus/Grafana.

TypeScript GraphQL Google Kubernetes Engine Pub Sub Vertex AI Pyarrow BigQuery Apache Airflow Prometheus Grafana

Senior Software Engineer

OpenX

Apr 2019 - Mar 2023 (3 years 11 months)

Developed the Hadron Service integrating the OpenRTB protocol for partner contextual segment targeting and scaled production traffic with throttling tests. Automated audience activation and ID export pipelines using Airflow batch jobs, integrated GraphQL/BigQuery for cross-team reporting via Looker Studio, and designed early Lookalike ML models with PySpark.

OpenRTB Protocol Apache Airflow BigQuery GraphQL Looker Studio Pyspark GDPR Compliance

Staff Software Engineer

TiVo

Feb 2018 - Mar 2019 (1 year 1 month)

Built a Danube-UDF pipeline combining batch and near real-time processing using Kafka and Spark with A/B testing. Migrated event-driven messaging from RabbitMQ to Kafka using Avro schema serialization to improve versioning/data compactness, and optimized KafkaProducer and consumer group rebalancing to enhance throughput and reduce latency.

Apache Kafka Apache Spark A B Testing Avro Serialization RabbitMQ Distributed Systems Performance Tuning

Principal Software Engineer

Rovi Corporation

Jul 2017 - Feb 2018 (7 months)

Diagnosed and resolved Out-Of-Memory issues in the Danube Resolver using VisualVM profiling and heap dump analysis, reducing memory footprint. Evaluated garbage collection algorithms to optimize backend performance and inform architectural decisions.

VisualVM Dump Analysis

Senior Software Engineer

Rovi Corporation

Jan 2014 - Jul 2017 (3 years 6 months)

Maintained the Danube Data Pipeline processing over 2 million records per hour. Administered MongoDB clusters to ensure high availability and reliability using advanced shell commands and aggregation frameworks.

MongoDB Administration MongoDB Aggregation Frameworks Data Pipelines High Availability Performance Monitoring