Vetted PyTorch Professionals

Abhay Murjani

Screened

Director-level Data Science Manager specializing in ML forecasting, experimentation, and MLOps

New York, NY6y exp

“Data/ML engineer with experience at American Express and Amazon, owning an end-to-end rewards redemption/liability ML pipeline (~200GB) with rigorous regulatory/audit validation and quarterly executive reporting. Also built web-scraped product datasets with anti-bot protections at a startup and helped modernize an authn/authz service using AWS, plus led early-stage migration work from an internal warehouse to GCP with CI/CD and cloud observability.”

“Built and productionized a multi-agent LLM analytics assistant at eBay that routes natural-language questions to retrieval or text-to-SQL, dynamically retrieves relevant schemas via a vector DB, and executes against a data warehouse. Drove a major quality lift (text-to-SQL accuracy 60%→85%) and materially reduced time engineers/PMs spent getting data insights through strong eval/monitoring, tracing, and reliability-focused design (schema retrieval, strict JSON outputs, retries/clarifications).”

“Backend/AI systems engineer who has shipped production LLM agents focused on prompt engineering, code generation, and incident-response automation. Stands out for combining strong agent orchestration and reliability engineering with measurable business impact, including 60-70% cost reductions, 45% lower monthly LLM spend, and a 5x increase in developer iteration speed.”

Alexander Smith

Screened

Junior Software Engineer and Data Scientist specializing in AI/ML systems

California, USA3y exp

“Built production-grade automation and ML/data pipelines at Dun & Bradstreet and ThreadNotion, spanning large-scale document classification, country risk report automation, and resilient Playwright testing for dynamic AI chat workflows. Particularly strong in turning brittle or ambiguous systems into reliable, observable, end-to-end automated platforms.”

“Robotics software engineer focused on ROS2 motion and calibration systems—built a trajectory generator/low-level controller using TOPPRA that improved robot motion speed by 11x while increasing accuracy. Experienced making high-frequency robot communication more real-time (core isolation) and shipping ROS2 modules via Docker-backed CI/CD, including serving as release manager coordinating reviews, release notes, and QA.”

Benjamin Kozel

Screened

Mid-level Machine Learning & Software Engineer specializing in RAG systems and ML infrastructure

Atlanta, GA4y exp

“Built and deployed an in-house RAG LLM system ("MONTY") using LLaMA 3B + FAISS to help teams quickly understand long internal/external specifications. Delivered usable production performance despite severe compute limits (single RTX 3080) by tuning retrieval/reranking and model choice, and is planning a LightRAG/knowledge-graph rewrite to improve accuracy and latency.”

Shuju Sun

Screened

Mid-Level Software Engineer specializing in real-time data pipelines and ML deployment

PA, USA4y exp

“Ticketmaster data engineer who built CDC-driven Kafka pipelines feeding Snowflake for analytics and data science teams. Hands-on in production operations—scaled Kafka during sudden playoff-driven transaction spikes and improved monitoring for preemptive scaling. Known for using small-batch experiments and quantitative metrics to align stakeholders and drive cost-saving architecture changes (e.g., buffering to reduce AWS Lambda invocation frequency).”

Ruturaj Ghatage

Screened

Junior Software Engineer specializing in full-stack, cloud infrastructure, and applied AI

Herndon, Virginia2y exp

“Master’s student at UC San Diego who built an LLM-powered healthcare chatbot for patient history-taking and sepsis-related output, using a Node.js backend integrated with FastAPI for RAG/LLM interactions and a Flutter client. Also has healthcare AI startup experience deploying on AWS (ECS/Terraform/Docker) and implementing Kubernetes autoscaling to improve efficiency and reduce costs, with strong iterative evaluation in collaboration with a physician.”

“Second most active contributor at Daice Labs, owning a production AI-powered software development collaboration platform’s end-to-end execution infrastructure (TypeScript/Next.js backend, Node.js CLI, shared libs). Built the full multi-agent pipeline (planning/codegen/summary), Supabase-backed context assembly and realtime state, Git/GitHub automation, and a provider-agnostic LLM abstraction with strict Zod validation and retries, backed by extensive tests and design specs.”

“ML/GenAI engineer with strong end-to-end production ownership across predictive ML, RAG systems, and LLM routing. They pair solid platform engineering skills with measurable business impact, including 15% churn reduction, 35% support ticket deflection, 45% GenAI cost savings, and a shared inference library that cut deployment time from weeks to days.”

“ML/AI engineer with hands-on ownership of production document intelligence and GenAI systems, spanning model experimentation, AWS deployment, monitoring, and iterative optimization. Stands out for turning document-heavy workflows into reliable, near real-time products with measurable gains in accuracy, latency, and manual-effort reduction, while also shipping citation-grounded RAG features that drove user trust and adoption.”

Harsh Sanas

Screened

Junior Full-Stack Engineer specializing in AI systems and cloud applications

San Francisco, CA2y exp

“Full-stack engineer with a strong applied AI bent who has built both a real-time EV charging platform and a production text-to-SQL system. Particularly compelling for teams needing someone who can bridge frontend, backend, infrastructure, and LLM evaluation/safety work, with experience shipping under early-stage ambiguity and integrating software with real-world hardware.”

“Backend engineer who built and iterated a secure, multi-tenant RAG system over a large document corpus, emphasizing strict RBAC/ACL isolation, hybrid retrieval (vector+keyword), reranking, and strong observability to balance relevance, latency, and cost. Also led production refactors/migrations using strangler + feature flags/dual writes and has experience catching subtle real-world failure modes (including in a sensor calibration optimization pipeline).”

Sarthak Gupta

Screened

Mid-level AI/ML Engineer specializing in LLMs, NLP, and real-time AI systems

New York, NY4y exp

“Backend engineer who built a real-time pipeline for recording, transcribing, and analyzing audio from 400+ news radio stations, scaling Whisper on an HPC cluster with 7 H100 GPUs. Has strong performance optimization experience (30% latency reduction via SQL/query design; 50% DB call reduction via Redis caching) and has implemented region-based data isolation and PII protections in a regulated environment (JP Morgan Chase).”

Robert Zhu

Screened

Mid-level Robotics & Systems Engineer specializing in GPU virtualization and autonomous manipulation

Toronto, Canada3y exp

“Robotics software engineer focused on vision-guided manipulation with a Franka Emika Panda on ROS 2, using an RGB-D camera and a trained YOLO-OBB detector to estimate 3D object poses for pick-and-place. Strong in perception-to-planning integration (TF2 alignment, message synchronization, MoveIt 2) and in debugging sim-to-real issues, including symmetry-aware orientation handling and grasp repeatability tuning.”

Darsh Sharma

Screened

Mid-level Software Engineer specializing in ML systems and microservices

Madison, WI2y exp

“Teradata Text Security intern who built a production LLM-powered planner agent that decomposes complex tasks into dependency-aware subtasks (DAG/topological graph) and executes them via a custom orchestrator with parallelism, status tracking, and error handling. Also contributed to an HR-facing internal document chatbot concept to streamline onboarding, showing cross-functional collaboration.”

“ML/AI engineer focused on production-grade model reliability: built a monitoring and validation framework to detect drift, trigger anomaly alerts/retraining, and maintain consistent performance for device intelligence workflows at scale. Strong MLOps background with Python pipelines, Docker/Kubernetes deployments, Airflow orchestration, and real-time monitoring dashboards; experienced partnering with product managers to deliver business-facing insights.”

“ML/AI engineer with hands-on ownership of production systems spanning classical ML fraud detection and GenAI agent workflows. At Fidelity, they built an end-to-end fraud platform that improved review queue Precision@K by 15-20% while reducing false positives 10-15%, and they also shipped RAG-based agent systems that cut manual workflow effort by 30-40%.”

“AI-native software engineer who uses agentic development as a core workflow, including a three-agent setup for planning, validation, and implementation. In their most recent role, they acted as the lead orchestrator for AI agents, with a strong emphasis on production safety, architectural control, and rigorous validation.”

“Robotics/ML engineer with hands-on experience building multimodal waypoint prediction for autonomous driving using CLIP + LidarCLIP embeddings and PyTorch, including nuScenes data pipelines and baseline modeling. Also built ROS 2 nodes for TurtleBot maze navigation with an image-classification pipeline, and has Caterpillar experience doing dSPACE HIL testing with MATLAB/Simulink plant models for engine software validation.”

“Applied Scientist with legal-tech experience who builds production LLM systems. Created and deployed Quibo AI, a LangGraph-based multi-agent pipeline that turns large markdown/Jupyter inputs into polished blogs and social posts, overcoming context limits via ChromaDB + HyDE RAG. Also built a large-scale iterative code-evolution workflow using multi-model orchestration (GPT/Claude/Gemini) with testing, debugging loops, and evaluation/observability practices.”

“AI/ML engineer who has shipped both a safety-critical mental health RAG chatbot (Mistral 7B + Pinecone) with automated faithfulness/toxicity monitoring and a deep Q-learning investment recommendation engine at Lincoln Financial Group. Strong in production MLOps and orchestration (AWS Lambda/CloudWatch/SageMaker, Docker, AKS) and in translating regulated-domain requirements (clinical reliability, fiduciary duty) into measurable model constraints and monitoring.”

“Built and shipped a production multi-agent, voice-based conversational assistant for older adults’ daily health management using Vertex AI, FastAPI, Firebase/Firestore, and Cloud Run, with a custom cross-session memory design to keep responses context-aware at low latency. Also partnered with caregivers/elderly users and health officials, translating needs into workflows and explaining HIV risk predictions with SHAP and dashboards.”

“AI engineer at Anote.ai who built and shipped a production multi-agent LangGraph/LangChain/Ray RAG platform for enterprise search and workflow automation, supporting 3 commercial products and 100+ developers. Drove measurable gains (30% accuracy improvement, lower latency) and improved reliability with Redis-based state checkpointing, message-queue synchronization, and Milvus retrieval optimizations, while partnering with PMs/clients to add transparency features like confidence scores and real-time logs.”