Vetted Classification Professionals

“Built and deployed a production generative AI chatbot at NVIDIA using LangChain + GPT-3 integrated with internal data sources, cutting response time nearly in half and improving CSAT by ~12 points. Also delivered LLM-driven QA tools by fine-tuning Hugging Face transformer models and deploying via an AWS-based pipeline (Lambda/Glue/S3) with orchestration (Airflow/Step Functions), CI/CD, Kubernetes, and monitoring (MLflow/Splunk/Power BI).”

“ML/NLP engineer focused on production GenAI and data linking systems: built a large-scale RAG pipeline over millions of support docs using LangChain/Pinecone and added a LangGraph-based validation layer to cut hallucinations ~40%. Also built scalable PySpark entity resolution (95%+ accuracy) and fine-tuned Sentence-BERT embeddings with contrastive learning for ~30% relevance lift, with strong CI/CD and observability practices (OpenTelemetry, Prometheus/Grafana).”

“Healthcare data platform builder with experience at Aetion delivering a rule-based EMR/EHR ingestion and validation framework that cut onboarding from 8–10 weeks to hours and unlocked $30M+ in revenue over ~3 years. Motivated to found an AI/agent-driven healthcare solution, with a specific interest in using PET scans, doctor notes, and treatment data with LLMs to help predict cancer progression and guide next-step treatments.”

“Internship experience building a production LLM-powered podcast operations agent that automated lead intake (HubSpot), guest research, scheduling (Calendly), meeting-summary evaluation (Gemini), and human approval via Slack bot—while retaining rejected candidates for future outreach. Also contributed to ideation of a multi-agent orchestration framework with parsing and task routing, and emphasized reliability via structured prompts, HITL feedback, and prompt-based test sets.”

“Built a production LLM-powered analyst assistant at Discern Security to speed up SOC investigations using a RAG pipeline over security vendor documentation (Python PDF ingestion, vector search). Demonstrates deep, security-critical LLM engineering: structure-aware chunking with custom table parsing, grounded/cited responses, prompt-injection defenses, and post-generation validation, validated via golden datasets and adversarial testing; tool is used daily by analysts.”

“Built and deployed a production healthcare medical review agent that automates call-transcript summarization and medication reconciliation using a hybrid deterministic + LangGraph-orchestrated LLM workflow. Demonstrates strong reliability engineering (guardrails, schema validation, confidence thresholds, golden/adversarial eval, Langfuse monitoring) in a regulated environment, delivering 60% lower latency and 70%+ efficiency gains while partnering closely with care managers and operations.”

“AI/LLM engineer at Capital One who deployed a production RAG-powered fraud analysis and document intelligence platform using LangChain, OpenAI, Pinecone, Kafka, and AWS. Focused on reliability in real-time investigations via hybrid retrieval, schema-validated outputs, and LLM verification loops, reporting review-time reduction from hours to minutes and ~99% fraud detection precision.”

“Built an "AskMyVideo" system that turns YouTube videos into queryable knowledge graphs by transcribing audio (Whisper), chunking and embedding content, and enabling traceable answers back to exact timestamps. Strong in entity resolution (rules + fuzzy matching + TF-IDF/cosine with PR-curve thresholding) and modern retrieval stacks (FAISS, hybrid dense/sparse, domain fine-tuning with ~12% precision gain), with a production mindset using Airflow/Prefect, Docker/FastAPI, and LangSmith/Prometheus/Grafana observability.”

“Built and shipped a production LLM-powered multi-agent RAG system to automate complex internal support workflows, integrating tool execution (SQL/APIs) with validation guardrails to reduce hallucinations. Optimized for real-world latency and cost via model routing, caching, and async parallel tool calls, and enforced reliability with CI-gated golden test sets derived from anonymized production queries.”

“ML/NLP practitioner working across healthcare and business/finance use cases: currently fine-tuning a domain-specific Llama 3.1 model for safe reasoning over EHRs/clinical notes using RAG + RL/DPO and RAGAS-based evaluation. Has built UMLS-driven entity normalization pipelines with quantified quality gains and developed embedding/vector-DB systems (FAISS) for semantic matching and forecasting/recommendation applications at Aurora AI and Banxico.”

“Backend/AI workflow engineer at a startup building production AI services for fashion workflows, including an AI-powered techpack generation API in Go (Gin) with MongoDB handling ~1k+ daily requests. Recently implementing an image-to-3D dress generation feature end-to-end, integrating a Python FastAPI AI service with ComfyUI + Hunyuan, with strong emphasis on async orchestration, webhooks, and observability (OpenTelemetry + SigNoz).”

“Walmart engineer who built and shipped a production LLM+RAG system to automate triage and analysis of computer support chats/tickets, producing grounded, schema-constrained JSON outputs for summaries, urgency, and routing recommendations. Emphasizes reliability (hallucination control, confidence thresholds, human-in-the-loop) and runs end-to-end pipelines with Airflow and AWS-native orchestration, plus rigorous evaluation and monitoring tied to business KPIs.”

“Built and deployed a production LLM/RAG fraud investigation system to replace manual investigator workflows, combining transaction data, historical cases, and policy documents with agent-style steps and LoRA fine-tuning. Demonstrates strong reliability engineering (grounding, citations, abstention paths), performance optimization (retrieval/indexing/caching), and end-to-end MLOps orchestration using Azure ML Pipelines/MLflow plus Kubernetes/Argo with canary and rollback deployments.”

“Built and shipped a production LLM-powered RAG assistant for enterprise internal document search (PDFs, knowledge bases, structured data), addressing real-world issues like noisy documents, hallucinations, and latency with grounded prompting, retrieval-confidence fallbacks, and performance optimizations. Also partnered with compliance and business teams at JPMc to deliver a solution aligned with regulatory constraints, supported by monitoring, feedback loops, and systematic evaluation.”

“Built a production assistive-vision iPhone app to help visually impaired users find grocery items, training a custom YOLO detector on 2,000+ self-collected/annotated images and deploying via CoreML with a cloud multimodal LLM for navigation instructions. Brings hands-on AWS serverless + ECS container deployment (CDK/GitHub Actions) and a disciplined approach to AI workflow reliability (state-machine design, offline evals, stress tests, logging/metrics), plus experience communicating model insights to non-technical stakeholders (MOTER Technologies).”