Vetted BERT Professionals

“Built a production SQL chatbot for access-log analytics that replaced manual custom report requests with natural-language querying, using LangGraph and a ChromaDB-backed RAG pipeline for grounded, consistent answers. Implemented a privacy-preserving design where the LLM never sees raw customer data (only query metadata) and has experience building multi-agent/tool-calling systems with LangGraph (DeepAgents), including solving sub-agent communication drift via self-reflection.”

“Built an AI lending assistant (RAG + DeBERTa) used by credit analysts to retrieve policies and past loan decisions, tackling real production issues like hallucinations, document quality, and sub-second latency. Deployed a modular, Dockerized AWS architecture (ECS/EMR + load balancer) with load testing, caching/precomputed embeddings, and CloudWatch monitoring, and used Airflow to automate scheduled data/embedding/vector DB refresh pipelines with retries and alerts.”

“Built and owned a hackathon project (Gritto) with a Python/FastAPI backend that routes user text through a sequence of Gemini agents to produce structured JSON outputs. Has hands-on production deployment experience using Docker/Docker Compose, GitHub Actions CI/CD, AWS App Runner, MongoDB, and secrets management (Doppler + migration to AWS Secrets Manager), plus implemented a chat-like experience via multiple HTTP requests when SSE wasn’t viable.”

“AI/ML Engineer who built a production RAG-based LLM system for insurance policy documents, turning thousands of messy PDFs into a searchable index using LangChain, Azure AI Search vectors, hybrid retrieval, and FastAPI. Strong focus on evaluation (MRR/precision@k/recall@k, REGAS) and performance optimization (vLLM), with prior clinical NLP experience using BERT-based NER validated on ground-truth datasets.”

“Built a secure, on-prem/private GPT assistant to replace manual SharePoint-style search across thousands of policies/SOPs/engineering docs, using a production RAG stack (LangChain/LangGraph, FAISS/Chroma, PyMuPDF+OCR, vLLM). Implemented layout-aware ingestion (including table-to-JSON) and a multi-agent retrieval/generation/verification workflow with strong observability and compliance guardrails, delivering ~70% reduction in search time.”

“ML/AI practitioner with hands-on experience deploying models to production and optimizing for low-latency inference using pruning/quantization, with deployments on AWS SageMaker and Azure ML. Has orchestrated end-to-end ML pipelines with Airflow and Kubeflow (ingestion through evaluation) and emphasizes reproducibility via containerization and version-controlled artifacts, while effectively partnering with non-technical stakeholders using dashboards and business-aligned metrics.”

“AI/ML engineer who has deployed transformer-based NLP systems to production via Python REST APIs and Kubernetes on AWS/Azure, with a strong focus on latency optimization (p95), reliability, and scalable orchestration. Demonstrates pragmatic model tradeoff decision-making and strong stakeholder collaboration—improving adoption by making outputs more actionable with summaries, extracted fields, and confidence indicators.”

“ML/NLP engineer with recent Scotiabank experience building production-grade indexing automation over large-scale emails and customer databases, combining LLM fine-tuning (Mistral, XLM-R) with fuzzy matching to exceed 95% accuracy under strict banking constraints. Also built a RAG-based chat agent using Gecko embeddings, Vertex AI Search, Gemini, and cross-encoder reranking, and delivered a text-to-SQL chatbot at SOTI through iterative fine-tuning and benchmark-driven experimentation.”

“Built and deployed an end-to-end product that brings a research-paper approach into production for large-scale time-series clustering, with attention to partitioning, latency, and scalability. Also designed a Python-based backend validation service (comparing outputs to database ground truths) and handled production reliability issues by reproducing dataset-specific crashes and hardening corner-case behavior with client-friendly errors.”

“GenAI engineer who has deployed production LLM/RAG chatbots for internal document search, focusing on reliability (hallucination reduction via prompt guardrails + retrieval filtering) and performance (latency improvements via caching). Experienced with LangChain/LangGraph orchestration for multi-step agent workflows and iterates using monitoring/logs and benchmark-driven evaluation while partnering closely with product and business teams.”

“GenAI/ML engineer with production experience at Cognizant and Ally Financial, building end-to-end LLM/RAG systems and ML pipelines. Delivered a domain chatbot trained from 90k tickets and 45k docs, improving intent accuracy (65%→83%), scaling to 800+ concurrent users with 99.2% uptime and sub-150ms latency, and driving +14% customer satisfaction. Strong in Azure ML + DevOps CI/CD, Dockerized deployments, and explainable/PII-safe modeling using SHAP/LIME to satisfy stakeholder trust and GDPR needs.”

“LLM/agent builder who has shipped production AI systems in the wellness space, including an LLM-powered food tracking product used by 5000+ users and a voice/call-routing onboarding workflow using LangGraph/LangChain with LiveKit and Twilio. Strong focus on practical reliability work: latency reduction, retrieval/embedding tuning, and CI-driven evaluation with simulations and metrics.”

“Backend engineer focused on AI-enabled systems, having built a production-style RAG pipeline (vector search + LLM) exposed via Python/Flask endpoints with strong observability and hallucination-reduction techniques. Demonstrates deep performance work in PostgreSQL/SQLAlchemy (5x faster analytics queries) and high-throughput optimization using Celery + Redis (800ms to 120ms latency, 3x throughput), plus schema-per-tenant multi-tenancy with tenant-aware middleware and logging.”

“Built and productionized Azure-based LLM/RAG systems for regulatory/compliance use cases, including automating analyst research and compliance report generation across large unstructured document sets. Demonstrates strong practical depth in hallucination mitigation, hybrid retrieval tuning (BM25 + embeddings), and production MLOps (Databricks, Cognitive Search, AKS, Airflow/MLflow), plus proven ability to deliver auditable, explainable solutions with non-technical compliance teams.”

“Built production LLM + hybrid RAG and multi-agent orchestration systems at Wells Fargo to automate complaint document/audio transcript understanding and categorization, addressing vocabulary drift via embedding + vector index updates instead of frequent retraining. Strong in LLM workflow reliability (testing/benchmarks/observability) and stakeholder-facing delivery with explainability (citations/SHAP-style justifications) and Tableau dashboards.”

“Senior data scientist with ~5 years’ experience building production ML/NLP systems in finance (Wells Fargo) and deep learning for sensor analytics in connected vehicles (Medtronic). Has delivered end-to-end platforms combining time-series forecasting with transformer-based NLP, including automated drift monitoring/retraining (MLflow + Airflow) and standardized Docker/CI/CD deployments; achieved a reported 22% precision improvement after domain fine-tuning.”

“BNY Mellon engineer who has built and operated production AI systems end-to-end: a LangChain/Pinecone RAG platform scaled via FastAPI + Kubernetes to 1000 RPM with 99.9% uptime, supported by monitoring and data-drift detection. Also deep in data/infra orchestration (Airflow, Dagster, Terraform on AWS/EMR/EC2), processing 500GB+ daily and delivering measurable reliability and performance gains, plus strong compliance-facing model explainability using SHAP and Tableau.”

“Built and deployed an LLM-powered claims-document summarization system (insurance domain) that cut agent review time from 4–5 minutes to under 2 minutes and saved 1,200+ hours per quarter. Hands-on across orchestration and production infrastructure (Airflow retraining DAGs, Kubernetes, SageMaker endpoints, FastAPI) and recent RAG workflows using n8n + Pinecone, with a strong focus on reliability, cost, and explainability for non-technical stakeholders.”

“Built and shipped production generative AI/RAG assistants in regulated financial contexts (S&P Global), automating compliance-oriented Q&A over earnings reports/filings with grounded answers and citations. Experienced across the full stack—AWS-based ingestion (PySpark/Glue), vector retrieval + LangChain agents, GPT-4/Claude model selection, and production reliability (monitoring, caching, retries) plus rigorous evaluation and regression testing.”

“AI/ML Engineer with 4+ years of experience (Capital One, Odin Technologies) and a master’s in Data Analytics (4.0 GPA) who has deployed LLM/RAG systems to production for compliance/risk and document review. Strong in orchestration and MLOps (Airflow, Kubernetes, MLflow, GitHub Actions) and in tackling real-world LLM constraints like latency, context limits, and data privacy, with measurable impact (20%+ manual review reduction; 33% faster release cycles).”

“Software engineer with Prudential Financial experience building enterprise Spring Boot microservices for policy/risk assessment, including integrating Python ML models via Flask and hardening services with resiliency patterns. Also led an AWS lift-and-shift modernization during an internship (EC2/ELB/Route53/Auto Scaling) and built a personal diffusion-model text-to-music project using BERT tokens mapped to Mel spectrograms.”

“Built a production, real-time clinical documentation system at HCA that converts doctor–patient conversations into structured clinical summaries using speech-to-text, LLM summarization, and RAG. Demonstrated measurable gains from medical-domain fine-tuning (clinical concept recall +18%, ROUGE-L 0.62 to 0.74) while meeting HIPAA constraints via PHI anonymization and encryption, and deployed via Docker/FastAPI with CI/CD and monitoring.”

“Healthcare-focused LLM engineer who deployed a production triage and clinical knowledge retrieval assistant using RAG and LangGraph-orchestrated multi-agent workflows. Emphasizes clinical safety and compliance with robust hallucination controls, HIPAA/PHI protections (tokenization, encryption, audit logging, zero-retention), and human-in-the-loop escalation; reports a 75% latency reduction in a healthcare agent system.”