AI in Production: Lessons Learned
summary
6sense consolidates legacy Hive Spark and HBase pipelines into SingleStore to serve tens of terabytes with 27 months retention and subsecond account queries, replacing thousands of batch jobs and cutting TCO while keeping per-rep UIs fast. Three concrete AI builds use OpenAI: an account AI Insights summary that scans 30–90 days (opps >1 year) and loads instantly, an email generator seeded from profile and behavior data, and Sherlock, an internal Slack Q&A backed by SingleStore vector search and RAG now rolling to customers.
AI-written summary · about
chapters (7)
- 0:00Introduction and company overview
- 3:57Data collection and ML platform
- 8:33Database consolidation with SingleStore
- 9:13AI Insights for account summaries
- 11:00AI-powered email writer
- 12:24Sherlock AI support bot
- 15:17Non-AI features and future enhancements
related talks
56mAI War Stories: Real world learnings delivering enterprise AI solutions
58mAI-Enhanced Big Data: Integrating Private LLMs and Vector Databases
33mEngineering AI Applications: A Conversation with AI Leaders
11mAI took over dev. Prod is next. Here's why it has to be
58mAI in Reality Fireside Chat: Enterprise AI & Open‑Source Innovation
27mAI as a Service Scaling AI with Ease and Innovation
36mBeyond the Hype
40mBuilding secure and efficient AI solutions in Azure with Partners