# Great Day Improvements × NeuralSeek — Customer Highlight > yETI — Secure, governed AI chat for HR and the Call Center Canonical URL: https://neuralseek.com/highlights/great-day-improvements Content-Type: text/markdown; charset=utf-8 NeuralSeek is building **yETI**, a ChatGPT-style chat experience for Great Day Improvements covering two lines of business — HR and the Call Center — with multi-LLM selection, secure authentication, Great Day's knowledge pre-loaded, and a Phase II path into the customer's Databricks datalake. Delivered end-to-end on Great Day's existing Azure environment. - Customer: **Great Day Improvements** - Industry: **Home Improvement** - Deployment: in Great Day's existing **Azure** tenant - Project name: **yETI** - Lines of business covered: **HR + Call Center** --- ## Scale of the customer - **~$1B+** annual revenue — estimated combined annual revenue across the Great Day family of brands. - **3,000+** employees — installers, sales, manufacturing, and corporate staff nationwide. - **Top 5** market position — one of the largest direct-to-homeowner home-improvement platforms in the U.S. (Patio Enclosures · Champion · Stanek). --- ## Voice of the client **Soo-Jin Behrstock** — Chief Technology Officer, Great Day Improvements LinkedIn: https://www.linkedin.com/in/soo-jin-behrstock-8198801/ As CTO of Great Day Improvements, Soo-Jin leads technology strategy across the family of brands and partners with NeuralSeek on the yETI rollout covering HR and the Call Center. ### Q&A **Q: What parts of NeuralSeek impress you?** A: What stands out is how hands-on and engaged the team is. They're not just pushing a product — they're in it with us. When we bring up a use case, they work through it with us and quickly turn it into something real we can see and react to. That ability to move from idea to proof of concept fast has made a big difference for us internally. **Q: Thoughts on NeuralSeek having an implementation arm and experience doing enterprise go-lives.** A: There were some bumps along the way, which is expected in something this complex. What stood out was how Lawrence and Jordan (PM) brought in the right people at the right time to work through issues quickly. The team's experience and ability to apply best practices made a difference in keeping things on track. Once development was complete, they stayed engaged through iterative testing across different scenarios, partnering closely with us to make sure everything worked as expected before going live. **Q: General feedback on our team.** A: Overall, it's been a strong partnership. The team is responsive, easy to work with, and clearly cares about getting this right. They're available when needed, they bring ideas to the table, and they collaborate well with our team. It feels like a working session, not a vendor relationship. --- ## About Great Day Improvements - Industry & sector: **Home Improvement** - Deployment: **Azure Cloud** — in Great Day's existing Azure tenant - Founded: **Est. 1966** — flagship Patio Enclosures brand founded - Headquarters: **Macedonia, OH** — with a national footprint - Ownership: **Private** — privately held, PE-backed - Brand portfolio: **Multi-brand** — Patio Enclosures · Champion · Stanek · more --- ## What's inside yETI ### 01 · Multi-LLM selection Users choose between multiple large language models inside yETI, matching the model to the task — concise answers, long-form reasoning, or domain-specific responses. Tags: Phase I · Model choice. ### 02 · Authenticated access Every user signs in before entering yETI. User identity flows through each session so answers, history, and uploaded documents stay scoped to the authenticated person. Tags: Phase I · Secure login. ### 03 · Document upload & analysis Users upload Word and PDF documents into the conversation and ask questions, summarize, or extract content inline — no separate analyst pipeline required. Tags: Phase I · Word / PDF. ### 04 · Chat history & memory Every user sees their previous questions in a running history and yETI remembers prior turns within a conversation, so follow-ups stay contextual instead of starting from zero. Tags: Phase I · Context-aware. ### 05 · Great Day knowledge preloaded Great Day Improvements' internal PDFs and Word documents are pre-indexed inside yETI, so users can ask natural-language questions about company policy, process, and product data directly. Tags: Phase I · RAG. ### 06 · Azure-native deployment yETI is deployed end-to-end inside Great Day's existing Azure environment — no new cloud footprint, no data leaving the customer's tenant. Tags: Phase I · In-tenant. ### 07 · Phase II — Databricks integration & advanced document ingestion The next phase extends yETI into Great Day's Databricks datalake with OAUTH or SAML single sign-on, plus richer document ingestion for CSV, Excel, PDF, and Word — so analysts can upload structured datasets and have the chatbot reason over them alongside the existing corpus. Tags: Phase II · Databricks · SSO. --- ## KPIs at a glance - **2** lines of business — HR + Call Center - **Multi-LLM** — model choice per query - **Azure** — in customer's existing tenant - **Phase II** — Databricks datalake next --- ## Related resources - Customer highlights index: https://neuralseek.com/Customer_Highlights - Platform overview: https://neuralseek.com/ - 118 AI Guardrails: https://neuralseek.com/guardrails - LLM Governance: https://neuralseek.com/llm-governance - Agent Governance: https://neuralseek.com/agent-governance - Book a working session with the team: https://calendar.app.google/VGE7TwcbKcaYsjRZ7