Architecture teardowns
& deep dives.
The engineering challenges behind production retrieval, agent workflows and applied machine learning. Explore the architecture, reported impact and companion evaluation examples.
Hybrid GraphRAG: Merging Neo4j with Vector Retrieval
A production RAG application combining graph relationships, vector search and multi-agent workflows.
Multi-Agent Orchestration for Autonomous Root-Cause Analysis
Retrieval and agent workflows that bring incident history into investigation and triage.
Recommendation Pipelines at Scale with Databricks & Spark
Production data and ML pipelines that turn customer interactions into personalized product recommendations.
Inventory Analytics: Turning Stockouts into Revenue Signals
A Streamlit dashboard that makes potential revenue loss from inventory issues easier to investigate.
SWYM · Production experienceMultiple Virtual Assistants in One Chatbot
A client chatbot bringing multiple virtual assistants into a single conversational interface with LUIS and Bot Framework.
Dhan AI · Production experienceHospital Help-Desk Conversations & Clinical NLP
A hospital help-desk chatbot built with Rasa, alongside ClinicalBERT adaptation work for project requirements.
Dhan AI · Production experienceAnomaly Detection: Reducing False Alerts in IT Support
Anomaly-detection models for IT support data, supported by internal ML and data workflows.
NLP for Support: Turning Ticket Text into Sentiment Signals
An NLP classification system that turns support-ticket text into sentiment signals.