Vetted Reinforcement Learning Professionals

“Built and deployed a production LLM-powered text-to-SQL/document intelligence chatbot on AWS that lets non-technical business users query complex enterprise databases in plain English. Demonstrates deep practical expertise in schema-aware prompting, embeddings-based schema retrieval, SQL safety/validation guardrails, and rigorous offline/online evaluation with human-in-the-loop approvals for risky queries.”

“Robotics software engineer with experience spanning safety-critical embedded medical hardware (low-cost neonatal baby warmer with PID temperature regulation) and advanced multi-robot planning research (belief-space planning with abstraction + MCTS to handle uncertainty). Strong ROS/ROS2 practitioner (Nav2/SLAM Toolbox/MoveIt) who builds custom packages (e.g., Insta360 panoramic imaging) and is hands-on debugging real robots from SLAM/frontier exploration to multi-robot collision avoidance and real-time performance.”

“Built a production-grade, semi-automated document recognition and classification system for large volumes of scanned PDFs, starting from little/no labeled data and handling highly variable scan quality. Deployed on AWS using SageMaker + Docker and orchestrated on EKS with a microservices design that scales CPU-heavy OCR separately from GPU inference, with strong reliability controls (validation, fallbacks, retries, readiness probes).”

“Built and deployed a production-grade, multi-agent Text-to-SQL assistant that lets non-technical stakeholders query large enterprise databases in natural language. Uses Pinecone-based schema retrieval + LLM reasoning (Gemini/Claude/GPT) with a dedicated validation agent (schema/syntax checks and safe dry runs) to reduce hallucinations and improve reliability, while optimizing latency and cost via async execution and embedding caching.”

“Graduate student taking advanced coursework in NLP, generative image modeling, and computer vision; built a PPO reinforcement-learning agent for a Super Mario platformer with careful reward shaping and metric-driven evaluation. In a recent internship designing credit risk models, created a 10-method feature-selection voting framework and achieved ~10% out-of-sample performance improvement while reducing features to mitigate overfitting.”

“Data scientist who has built production LLM agents (GPT-4o + LangChain + RAG) to automate analyst-style ad hoc CSV analysis with guardrails and GPT-as-a-judge evaluation. Also delivered an explainable healthcare NLP system for ICD code classification by collaborating closely with clinicians, using a hybrid rule-based decision tree + BERT model to reach 97% accuracy and cut manual review time.”

“AI/ML engineer who has shipped production systems across computer vision and conversational agents: built a YOLOv8-based wheel fitment pipeline at a Techstars-backed automotive startup, focusing on sub-second latency, monitoring, and robust fallback mechanisms that drove 2–3x page view growth and +5–6k users. Also built a voice-based interview platform orchestrating Deepgram + GPT-4 Mini + OpenAI TTS with FSM-driven reliability, and has hands-on RAG experience (LangChain, hybrid retrieval, cross-encoder reranking, custom pseudo-query generation).”

“Built and deployed production RAG-based document search/Q&A systems (DocChat and an internship marketing RAG), using a React + FastAPI stack on GCP with docs stored in GCP buckets and retrieval via embeddings/vector DB. Emphasizes cost/performance tradeoffs (reported ~40% cost reduction) and ships via Docker (Railway), with load/API testing using JMeter and Swagger; regularly collaborates with a CEO stakeholder to iterate and push changes to production.”

“Robotics software engineer with hands-on SLAM, ROS2, and distributed multi-robot systems experience. Improved MAST3R-SLAM loop-closure place recognition by changing the ASMK/ASMKS retrieval similarity metric (L2→L1) and validated on 9 TUM sequences, keeping near real-time performance despite a 25–30% retrieval cost increase. Also tuned MoveIt motion planning for a 6-DOF arm (12% higher maze completion rate) and built MQTT mesh communications for ESP32-based AMRs, using Gazebo+Docker and CI-style automation for reproducible testing and deployment.”

“Robotics/SLAM-focused engineer who worked on RT-Appearance mapping using NetVLAD, replacing traditional CV feature extraction with a deep learning approach to improve loop closure in repetitive green environments. Has hands-on ROS1/ROS2 experience (including bridging), point-cloud alignment with G-ICP for sensor-parameter matching, and Gazebo+Docker simulation testing for motion planning/perception.”

“LLM engineer with production experience building end-to-end document processing workflows that unify layout analysis, OCR, and downstream LLM reasoning. Has implemented reliability features (retries, robust error handling, OpenTelemetry logging) and built agentic systems using LangChain/CrewAI, including a student research-paper assistant, while collaborating closely with PMs and non-technical end users to reduce technical debt and simplify architectures.”

“Robotics software engineer who built a ROS 2 Humble autonomous hospital-equipment detection/localization robot end-to-end in Gazebo (custom worlds/models, Nav2 waypoint navigation, YOLOv8n perception, TF2-based depth fusion) and solved real-time integration issues via multithreading and QoS tuning. Also implemented and tuned an MPPI controller to enable smooth reverse parking on an OpenPodCarV2 platform, including real-world reverse engineering and hardware/software debugging.”

“Built and deployed a production-grade LLM agent for credit management and accounts receivable automation, integrating ERP/MySQL data via a RAG pipeline and exposing services through FastAPI with Pydantic-validated outputs on AWS Bedrock. Emphasizes reliability and compliance for financial operations using schema validation and human-in-the-loop review, reporting ~32% reduction in manual work and ~41% improvement in response time/reliability.”

“Robotics engineer focused on simulation, modeling, and control, with hands-on sim-to-real experience from a soft, foldable “grasshopper” robot where friction/contact physics and servo dynamics drove real-world performance gaps. Built a ROS 2 voice-operated TurtleBot system integrating YOLOv5 + stereo depth for object picking with an attached arm, and debugged AMCL/SLAM to cut localization error from 10–13 cm to ~5 cm. Currently developing a quadruped in MuJoCo with a 3-layer control stack (RL + MPC + PD) and an RL training pipeline in JAX ahead of hardware.”

“Robotics software engineer focused on UAV multi-agent collision avoidance, combining Behavior Trees with MPC and validating in a ROS-based simulator via PX4 hardware-in-the-loop. Has hands-on experience debugging real-time estimation/control issues (EKF noise/arming failures), building distributed ROS service architectures, and smoothing MPC trajectories to meet kinematic and FAA right-of-way constraints.”

“AI/LLM engineer who built and deployed a production healthcare RAG chatbot ("DoctorBot") with strict medical safety guardrails, an 85% confidence-gated verification layer, and latency optimizations that cut responses from ~8s to ~2–3s. Also worked on finflow.ai to generate finance/banking test cases from BRDs, collaborating closely with non-technical domain stakeholders, and has hands-on orchestration experience with LangChain/LangGraph and agentic evaluation/monitoring practices.”

“PhD in computer engineering who has built production-oriented ML/NLP systems for space-weather prediction using Spark-based ETL on noisy satellite sensor logs. Strong in entity resolution and semantic search—fine-tuned E5 embeddings with contrastive learning and deployed to Pinecone, improving top-5 retrieval precision by 25%—and emphasizes scalable, observable pipelines with Airflow, Docker, and CI/CD.”

“Robotics engineer leading development of a Physical Reservoir Computing controller for a pneumatic soft robotic arm, owning everything from automated data collection and leak-testing automation to hardware design/manufacturing and cross-lab integration with Virginia Tech. Built ROS 2/DDS-based multi-robot systems integrating OptiTrack, a lab quadruped, and a UR5e, and pairs simulation (Gazebo/MuJoCo) + PPO RL training with production-ready tooling (Docker, CI/CD, Flask dashboards, RAG chatbot portfolio).”

“Robotics software candidate who built a Q-learning smart delivery drone navigation system, focusing on 2D path planning with dynamic obstacle avoidance using reward shaping and real-time sensor feedback. Actively learning ROS 2 by building Python simulation projects with publisher/subscriber patterns and has experience coordinating multi-agent drone simulations via message passing.”