Vetted Amazon SageMaker Professionals

“Built and productionized a real-time enterprise RAG pipeline to improve factual accuracy and reduce LLM hallucinations by grounding responses in constantly changing internal knowledge bases (policies, manuals, FAQs). Experienced in orchestrating end-to-end ML workflows (Airflow/Kubernetes), handling messy multi-format data with schema enforcement (Pydantic/Hydra), and maintaining freshness via streaming incremental embeddings plus batch refresh. Also delivers applied ML solutions with non-technical teams (marketing/CRM) for segmentation and personalized engagement.”

“Built and shipped an LLM-powered data quality assistant that generates maintainable validation checks from metadata while executing validations via Great Expectations, exposed through FastAPI and integrated into Airflow-managed pipelines. Emphasizes production reliability (structured outputs, guardrails, monitoring, versioning, human review) and works closely with compliance/operations teams to deliver clear, auditable, user-friendly AI outputs.”

“Former Amazon intern who owned an end-to-end computer vision system to detect package anomalies in fulfillment centers, from data collection/labeling to production deployment on AWS (EC2/S3) with a Streamlit live-monitoring dashboard. Also has ML-in-production experience deploying and updating a recommendation model on Kubernetes (Minikube) with CI/CD via GitHub Actions, plus prior SDE experience with Jenkins-based pipelines and on-prem to AWS migration work using Glue.”

“LLM/Conversational AI engineer who built a production multi-turn dialogue system using LoRA fine-tuning on LLaMA, cutting training compute/memory by 90%+ while maintaining low-latency inference via quantization and streaming generation. Experienced in orchestrating end-to-end ML workflows with Prefect/Airflow/Kubeflow (including hyperparameter sweeps and W&B tracking) and improving agent reliability through benchmark-driven testing, shadow-mode rollouts, and stakeholder-informed guardrails.”

“Engineering leader who modernized and unified a fragmented product suite at Milestone via a multi-year cloud-native roadmap, delivering an MVP in three quarters and boosting team velocity by 40% through cross-functional squads. At Prometheum, led a trust-building hybrid architecture (AWS control plane + customer-hosted data plane) using Kubernetes to ensure sensitive enterprise data never left customer networks while remaining cloud-agnostic across providers.”

“Built and shipped a production LLM-powered medical scribe that generates structured clinical visit summaries using RAG, strict JSON schemas, and post-generation validation to reduce hallucinations. Experienced in making LLM workflows deterministic and observable (structured logging/metrics/tracing) and in evaluation-driven iteration with metrics like schema pass rate and edit rate; collaborated closely with clinicians and policy stakeholders at Scale AI to drive adoption.”

“DevOps/SRE focused on cloud-based distributed systems, with strong hands-on Kubernetes production experience (microservices deployments, Helm, probes, resource tuning, CI/CD and Docker build standardization). Demonstrated end-to-end troubleshooting across application, infrastructure, and networking layers—e.g., isolating degraded storage via node disk I/O metrics and restoring performance by draining the node and replacing the volume. Builds Python automation for operational reliability, including scheduled Kubernetes secrets rotation integrated with an external secret manager.”

“Built and shipped a real-time AI phone agent for small businesses that handles bookings/FAQs/messages using streaming ASR, an LLM with tool-calling, and TTS; deployed to production for multiple paying customers. Demonstrates strong applied LLM reliability practices (tool-first grounding, retrieval, hard-negative testing, and production monitoring) and experience orchestrating multi-step AI workflows with Airflow, Prefect, and AWS Step Functions.”

“Player-coach backend leader at PostLo who led a major backend architecture upgrade to enable AI-driven features by separating transactional systems from AI workloads (vector embeddings/image validation) and adding async processing for heavy jobs. Also owned production reliability improvements (query/index optimization, workload isolation, monitoring and load testing) and translated an ambiguous retention goal into a shipped cashback rewards feature with auditable transactions.”