Vetted Prometheus Professionals

“End-to-end product engineer who built and shipped Shopify Magic, an LLM-powered product-description generator on Amazon Bedrock with RAG over a tenant-isolated vector database, achieving 50% faster content creation, sub-2s latency, and 70%+ merchant adoption. Also led a Flexport migration from a monolithic Rails app to microservices using feature flags and parallel runs, delivering zero downtime and a 60% improvement in development speed.”

YAKKALI PAVAN

Screened

Mid-level Machine Learning & Generative AI Engineer specializing in NLP, CV, and RAG systems

USA6y exp

“Built and deployed a production LLM-powered RAG document intelligence system used by non-technical enterprise stakeholders, cutting document search time by 40%+ while improving answer consistency. Demonstrates strong MLOps/data workflow orchestration (Airflow, AWS Step Functions, managed schedulers across GCP/Azure) and a metrics-driven approach to reliability, evaluation, and cost/latency optimization with guardrails and observability.”

“Fintech engineer with high-ownership startup experience at Column, where they led a Go rewrite of a core banking API that cut transaction errors by 98% and increased throughput by 40%. Brings a rare mix of backend systems depth in compliance-sensitive money movement infrastructure and frontend platform experience building React design systems, with earlier experience at Meta on recommendation APIs and News Feed.”

“Senior engineering leader and architect with 15+ years of experience building organizations and cloud-native platforms across cybersecurity, healthcare, fintech, and enterprise software. At A10 Networks, he served as the principal architect while managing 30+ engineers, delivering a multi-tenant SaaS platform and AI-powered DDoS product in under a year, then migrating legacy products and enterprise customers onto the new foundation.”

“Data science/NLP practitioner with experience at NVIDIA and Microsoft building production-grade NLP and data-linking systems. Has delivered high-performing pipelines (e.g., F1 0.92) and large-scale entity resolution (F1 0.89), plus semantic search using embeddings and Pinecone with ~30–40% relevance gains, backed by rigorous validation (A/B tests, ROUGE, MRR) and strong MLOps/workflow tooling (Airflow, Databricks, FastAPI, MLflow, Prometheus/ELK).”

“Engineering leader in cloud security at SysTech with player-coach experience spanning cross-team data/ownership standardization and reporting platform user-journey improvements. Stays technically deep through observability (SLA/SLOs, dashboards, alerting), rigorous code reviews (including AI-assisted coding), and end-to-end incident ownership in IAM/agentless cloud event collection. Targeting $270K–$300K base plus bonus/equity.”

Sandeep Rohilla

Screened

Principal Backend/Platform Engineer specializing in GenAI agent orchestration and LLM pipelines

San Francisco, CA19y exp

“LLM-focused engineer/sales-engineering profile with hands-on experience productionizing complex systems: scalable distributed architecture, multi-tenant monitoring, canary/shadow rollouts, and robust fallback strategies. Demonstrated real-time troubleshooting depth (p99 latency spikes traced to DB connection limits causing retry storms) and strong developer-facing communication via RAG workshops and live, customer-specific demos that helped close deals quickly.”

“Built and shipped production AI systems spanning customer support automation at Uber, privacy-preserving federated health modeling on iOS, and an open-source semantic search layer for Postgres. Stands out for combining strong LLM/product instincts with rigorous eval design, measurable production impact, and zero-to-one execution across backend, mobile, and developer infrastructure.”

“Built and owned end-to-end deployment and AI support workflows spanning CI/CD, Kubernetes, Terraform, and LLM/RAG systems. Stands out for combining DevOps delivery with production AI operations, including secure tool-calling, incident debugging, retrieval quality controls, and validation-first document ingestion for messy real-world inputs.”