Vetted Data Scientists

“Interned at Backblaze building production AI systems for incident response and security operations, including an internal LLM-powered incident triage assistant that used Snowflake + RAG over historical tickets/postmortems and delivered results via Slack and a web UI. Emphasizes reliability (PII filtering, grounding, schema validation, fallbacks) and rigorous evaluation/observability (offline replay, partial rollouts, time-to-first-action metrics, Prometheus/Grafana).”

“Built and shipped a production agentic LLM tool that helps internal teams update technical product whitepapers using plain-language edit requests, with strong guardrails (citations, verification, refusal/clarify flows) to reduce hallucinations and maintain compliance. Experienced taking LLM workflows from rapid LangChain prototypes to more predictable, debuggable LangGraph agent graphs, and orchestrating end-to-end ingestion/embedding/indexing/eval/deploy pipelines with Kubeflow.”

“Backend/data engineer focused on production ML/LLM systems: built a real-time FastAPI inference API on Kubernetes with strong reliability patterns (timeouts, idempotent retries, centralized error handling). Delivered AWS platforms using EKS + Lambda with GitHub Actions/Helm CI/CD and built Glue-based ETL from S3/Kafka into Snowflake with schema evolution and data-quality controls; also modernized legacy analytics/recommendation workflows into Python services with safe, feature-flagged cutovers.”

“Built and deployed a production LLM-powered clinical support assistant at BJC HealthCare (RAG + transformer) to answer patient questions, summarize clinical notes, and support appointment workflows. Implemented PHI-safe data pipelines (Spark/Hadoop/Kafka) with automated scrubbing, dataset versioning, and audit logs, and runs the system on Docker/Kubernetes with Pinecone vector search while partnering closely with clinical operations staff.”

“LLM/RAG practitioner who built an AWS-based enterprise document search and summarization platform with RBAC and scaled it to 10K+ users, solving relevance issues via contextual chunking and hybrid retrieval. Also designed agentic workflows for a telecom forecast-validation use case using sub-agents, tool APIs, and strict context management, and has proven pre-sales influence (supported a $300K manufacturing deal with a roadmap-driven pitch).”

“AI/LLM engineer who built and productionized an internal RAG-based knowledge system that ingests diverse sources (PDFs, Markdown, Slack), scaled retrieval with distributed FAISS and parallel ingestion, and reduced hallucinations via re-ranking, grounding prompts, and post-generation validation. Also has hands-on orchestration experience with Airflow and Kubernetes for reliable ETL/model pipelines, monitoring, and staged rollouts; reports ~15% accuracy improvement and adoption as the primary internal knowledge tool.”

“Built and deployed a healthcare NLP application that used an LLM-style physician interface feeding a random forest model to predict treatment plans for hard-to-triage patient subgroups, backed by a Databricks medallion pipeline and heavy feature engineering to address missing/low-integrity data across ~50K patients. Also delivered an earlier Microsoft AI Builder automation that improved transportation bill payment workflows by training non-technical payroll/procurement teams to use automated outstanding-payables reporting.”

“Built and deployed a production LLM-powered RAG document intelligence system used by non-technical enterprise stakeholders, cutting document search time by 40%+ while improving answer consistency. Demonstrates strong MLOps/data workflow orchestration (Airflow, AWS Step Functions, managed schedulers across GCP/Azure) and a metrics-driven approach to reliability, evaluation, and cost/latency optimization with guardrails and observability.”

“Data science/NLP practitioner with experience at NVIDIA and Microsoft building production-grade NLP and data-linking systems. Has delivered high-performing pipelines (e.g., F1 0.92) and large-scale entity resolution (F1 0.89), plus semantic search using embeddings and Pinecone with ~30–40% relevance gains, backed by rigorous validation (A/B tests, ROUGE, MRR) and strong MLOps/workflow tooling (Airflow, Databricks, FastAPI, MLflow, Prometheus/ELK).”

“ML/NLP practitioner with a master’s thesis focused on domain-adaptive knowledge distillation for LLMs (LLaMA2/sheared LLaMA), showing improved perplexity and ROUGE-L on biomedical data. Also built real-world data linking and search systems: integrated ClinicalTrials.gov with FAERS using fuzzy matching + embeddings, and delivered an LLM-powered FAQ recommender at Hyperledger using sentence-transformers, FAISS, and fine-tuning to mitigate embedding drift.”

“Full-stack engineer in the investment tracking/financial reporting space who built an automated reporting dashboard and compliance/reporting pipeline end-to-end using Next.js (App Router, server/client components), REST, and Postgres. Demonstrated measurable performance wins (~30% faster loads) through caching and query optimization, and built durable orchestrated workflows in n8n with retries, idempotency, and reconciliation checks.”