Vetted Natural Language Processing Professionals

“Built a production AI-powered workflow automation system at Netflix that integrated OpenAI and LangChain with FastAPI services on AWS, cutting roughly 320 hours of manual operational effort. Brings a mix of full-stack product development and practical AI systems experience, with strong attention to reliability, maintainability, and non-technical user adoption.”

“Robotics software engineer from Ati Motors who brought a Linear MPC approach (based on Kuhne et al.) into production, rebuilding parts of the planning stack to eliminate oscillations and safely double AMR speed from 0.8 m/s to 1.6 m/s. Also delivered an end-to-end point-cloud detection pipeline (PointPillars) including synthetic data generation in Isaac Sim and TensorRT deployment for real-time human/trolley detection, with a strong focus on production reliability via iterative hardening and nightly SIL.”

“Built and deployed a production LLM-powered risk assistant for KPMG and Freddie Mac that lets analysts query a confidential Neo4j risk graph in natural language (no Cypher), turning multi-day analysis into minutes with traceable, cited answers. Implemented rigorous guardrails, deterministic verification, RBAC/security controls, and a full eval/observability stack, cutting query error rate by ~50% and iterating through weekly UAT with non-technical risk analysts.”

Dheeraj Kumar

Screened

Intern Data Scientist specializing in marketing analytics and data engineering

Tucson, Arizona2y exp

“AI/LLM practitioner with internships at Dell Technologies and Roche who built and deployed a healthcare-focused "Doctor LLM" by fine-tuning Meta Llama 3.2 on healthcaremagic.json, emphasizing safety guardrails to prevent harmful medical advice. Experienced in productionizing AI workflows with monitoring, testing, and orchestration (Airflow, Kubernetes), and in delivering AI-agent-driven competitive landscape insights to non-technical business stakeholders.”

“Built GenMedX, a multi-module clinical AI system for emergency department decision support spanning triage prediction, diagnosis, medication Q&A, and visit summarization. Stands out for combining medical LLM fine-tuning, RAG, and rigorous evaluation/monitoring to drive a major triage recall improvement from 38.5% to 76.6%, with a strong focus on safety, edge-case detection, and production reliability.”

Meredith Ma

Screened

Entry-level AI/ML Software Engineer specializing in generative AI and computer vision

Pittsburgh, PA2y exp

“Built and owned a production RAG coding assistant at Magna International used by 200 engineers, with hands-on work across React/TypeScript, retrieval infrastructure, and Postgres observability. Also brings an unusual blend of product UX thinking from AR game onboarding work, showing strength in both technical systems reliability and user activation.”

“Senior product executive with leadership experience in both insurance and legal tech, including a 0-to-1 AI/ML visual intelligence claims platform at Solera that generated $15M ARR in year one. Stands out for combining enterprise product strategy, cross-functional execution, ML fluency, and financially disciplined product leadership, plus a track record of UX-led growth such as a 300% user-base increase at Epiq.”

“CMU IoT coursework project builder who implemented an end-to-end TinyML gesture recognition system on a Particle Photon + ADXL345, streaming data via MQTT/Node-RED to a real-time Node.js frontend and deploying a quantized logistic regression model on-device. Also explored multi-drone coordination, implementing leader-follower offset control and a pivot/arc turning strategy to avoid collisions, and brings practical Docker/Kubernetes plus CI/CD workflow experience from internships.”

“AWS engineer who productionized an internal ML-driven data pipeline from a notebook prototype into a scalable, observable Python service (schema validation, deduplication, idempotency, safe retries, versioned transforms, CloudWatch alarms), reducing manual effort and improving data accuracy/trust. Experienced diagnosing workflow issues in real time (e.g., upstream schema changes) and partnering with account managers/support to unblock adoption of seller-facing Marketplace features by demonstrating reliability with concrete metrics.”

“Built a production LLM-driven archival assistant that turns large, low-quality scanned handwritten files (120+ pages) into structured datasets, overcoming context-window and hierarchy challenges with a two-phase LLM + rules pipeline and reaching 98.1% accuracy (Gemini-2.5 Flash). Also orchestrated a large human-in-the-loop effort with 78 archivists, producing 2,400 high-quality annotations in 4 days via detailed rubrics and support.”

“Senior engineering leader in the Bay Area with experience spanning VMware, Hortonworks/Cloudera, Barracuda, and Palo Alto Networks, including leading open-source work (Apache Knox) and architecting large-scale security platforms. Has driven disaster recovery and cloud security products, designed Python microservices for Microsoft 365 security, and scaled teams (3x) while formalizing enterprise readiness practices with automated documentation using Notebook LLM.”

“Current financial data analyst at Maximus with hands-on experience building SQL and Python reporting pipelines for expenditure, refund, and retention analysis. Stands out for turning messy multi-source financial data into trusted dashboards, automating reporting to save 15+ hours weekly, and driving measurable reductions in refund rates through cohort-based analysis.”

“Built multiple AI-powered products from scratch, including ConnectAbility, an accessibility tool combining computer vision and LLMs to describe visual content for users with disabilities, and SpamBack!, a macOS app that detects scam texts and auto-generates responses. Stands out for full-stack/backend ownership of applied AI systems, especially around async workflows, inference performance, and reliability safeguards.”

“AI/LLM-focused engineer with hands-on experience building RAG pipelines, prompt engineering workflows, and multi-agent systems using tools like LangChain. Stands out for combining AI-assisted development with production-grade validation and for leading the architecture/orchestration of agent-based recommendation systems that improved response time, accuracy, and scalability.”

“Backend/data engineer with Optum healthcare claims domain experience building high-reliability Python microservices (FastAPI/Kafka/Postgres) and AWS data platforms (EKS, Glue, Redshift). Demonstrated strong production ownership: fixed duplicate Kafka processing via transactional outbox/idempotency, scaled to millions of daily events, and delivered major SQL performance gains (40+ min to <5 min, ~60% CPU reduction). Seeking remote-only work; targets $130k base.”

“Built production LLM systems focused on reliability and safety, including a plain-English deployment tool that generates validated plans and provisions to Kubernetes while preventing unsafe actions via schema enforcement and plan/execute separation. Also created multi-LLM workflows (LangGraph) and stakeholder-friendly demos at Bosch, including a PyQt/FastAPI/CUDA app comparing SAM2 vs SAMWISE for on-device object detection with intuitive UX for business users.”

“Backend/data engineer who has owned production Python APIs and high-throughput async workflows on AWS (FastAPI, Docker, ECS/EKS/Lambda) with mature reliability practices like idempotency, bounded retries, circuit breakers, and strong observability. Also built AWS Glue ETL into an S3/Redshift lakehouse and modernized legacy batch systems via parallel-run parity testing and feature-flagged migrations, including a SQL tuning win cutting a multi-minute query to under 10 seconds.”

“Machine learning software engineer intern experience at Amazon, where they built a production testing framework to inject frames/videos onto devices to measure embedded CV model inference and ensure broad model compatibility via automatic NNA metadata handling. Also built side projects spanning LLM/RAG orchestration (LangChain/LangGraph with reranking and citations) and applied CV/healthcare work (nail disease detection, medical retrieval chatbot).”

Swarnabha Roy

Screened

Mid-level Robotics & Computer Vision Engineer specializing in autonomous systems and edge AI

College Station, TX6y exp

“Robotics/perception researcher (MVOS Lab, South Dakota State University) who built an end-to-end multimodal RGB-D + LiDAR pipeline for autonomous greenhouse harvesting and 3D plant phenotyping. Demonstrated strong production ownership by diagnosing motion blur with ROS-bag + OpenCV metrics and shipping an edge-deployed, scan-quality-aware workflow that boosted barcode read rate to 98% and supported ~70% autonomous pepper detection/harvesting accuracy.”

Daniel Luzzatto

Screened

Junior Machine Learning Engineer specializing in LLMs, computer vision, and robotics

Tirat Carmel, Israel1y exp

“Built and deployed an agentic, multimodal LLM system that automates privacy redaction pipelines (audio/video/tabular) using LangChain orchestration and a closed-loop self-correction design. Personally implemented and performance-optimized core CV tooling (face blurring with tracking/Kalman filter) achieving >100 FPS on CPU, and validated reliability with golden-dataset benchmarking across 100+ privacy intents and measurable redaction metrics.”

Daksh Jain

Screened

Mid-Level Software Development Engineer specializing in AWS streaming media platforms

Portland, OR3y exp

“Full-stack engineer with hands-on Next.js App Router + TypeScript experience (built a RateMyProfessor-style platform end-to-end using RSC, dynamic routing, debounced search, and cache invalidation). Also has AWS backend depth—built a Step Functions-based wave rollout/feature access control framework for MediaPackage V2 with idempotency, retries, rollback, and ongoing correctness reconciliation.”

Yash Jajoo

Screened

Senior Software Engineer specializing in AI and FinTech platforms

New York City, NY8y exp

“Built a production LLM pipeline at Walter AI that scans massive user inboxes, identifies financial newsletters, and extracts trading strategies into structured JSON for downstream paper-trading workflows. Stands out for combining agent architecture with strong production discipline—cutting scan time from 20 to 5 minutes, reducing LLM costs by 90%, and achieving 3-second P99 latency while handling messy, inconsistent email data at scale.”

“Backend engineer with end-to-end ownership of scientific and AI-powered systems, including neuron imaging pipelines at Monell Chemical Senses Center and an LLM-based structured information extraction platform for Wharton and PSG. Stands out for turning messy, compute-heavy workflows into reliable production backends with measurable impact, including saving researchers over 50 hours per week.”