Larry Honrada

Senior AI Engineer with 10+ years building production-grade machine learning systems and distributed data platforms. Focused on LLM-powered applications, retrieval systems, and low-latency inference. Proven track record delivering systems that process millions of documents, reduce processing time from hours to seconds, and operate at scale under real-world workloads.

Senior AI Engineer

Luxoft,

Generative AI Document Intelligence Platform (RAG)

Dec 2022 – Mar 2026

• Led development of a production-grade RAG platform enabling natural language querying across millions of insurance documents (policies, contracts, structured records)
• Reduced document analysis time from hours to seconds by automating ingestion, semantic retrieval, and LLM-based answer generation
• Designed and implemented end-to-end pipeline: document ingestion, parsing, chunking, embedding, vector indexing, retrieval, and LLM inference
• Built scalable data pipelines using Python and Apache Airflow to process and normalize large volumes of unstructured and structured data
• Engineered embedding and indexing workflows using vector databases (FAISS / Pinecone / OpenSearch) to support semantic search over 1M+ document vectors
• Achieved ~38ms query latency by optimizing vector search and retrieval pipelines for high-throughput, low-latency workloads
• Implemented hybrid retrieval (vector + keyword/BM25) and reranking strategies to improve search relevance and contextual accuracy
• Developed LLM-powered inference pipelines using OpenAI / Azure OpenAI, enabling contextual Q&A and document understanding
• Built backend services and REST APIs (FastAPI, microservices architecture) to support real-time querying, summarization, and search endpoints
• Designed and deployed cloud-native infrastructure across AWS and Azure using Docker and Kubernetes (EKS/AKS)
• Implemented MLOps workflows including CI/CD pipelines, model versioning (MLflow), monitoring, and retraining pipelines
• Integrated observability and monitoring systems (Prometheus, Grafana, logging pipelines) to ensure system reliability and performance
• Collaborated with frontend teams to support chat-based UI for analyst interaction with enterprise datasets
• Owned system architecture and delivery, leading a team of 4–6 engineers and coordinating across ML, backend, and data engineering functions

ML Platform Engineer

Dataminr

Real-Time Event Detection ML Platform (AI Infrastructure Layer)

Aug 2017 – Nov 2022

• Built distributed ML infrastructure powering real-time event detection across global data streams
• Designed streaming pipelines handling billions of daily inputs (text, image, sensor data)
• Improved system throughput by ~30% through pipeline optimization and system tuning
• Reduced model deployment time from weeks to under one week via automation and CI/CD
• Developed feature pipelines for real-time and batch ML workflows
• Built model serving systems enabling low-latency inference at scale
• Implemented monitoring and alerting (Prometheus, Grafana, ELK), reducing incident detection time by ~50%

AI Engineer

Accenture

Telecom Customer Churn Prediction System

Feb 2013 – Jul 2017

• Built ML models (logistic regression, random forest, gradient boosting) for churn prediction
• Designed pipelines processing large-scale CRM and billing datasets
• Generated risk scores for millions of users to support retention strategies
• Improved model performance through feature engineering and evaluation (ROC-AUC, precision/recall)
• Deployed batch scoring systems for enterprise ML workloads

Bachelor’s Degree in Computer Science

August 2008 – Apr 2013

University of Philippines