Buildingwhat's next.
Where possibilities are turned into value.
At Winter, we believe innovation happens when human will meets machine speed. 25 Days of AI is our showcase series that brings this belief to life by giving emerging Forward Deployed Engineers the opportunity to take on real enterprise challenges and build AI proof-of-concepts. Working with frontier technology and partner platforms, the next generation of the Winter Squad moves from problem discovery to working solutions in an environment where attempting and learning matter as much as the outcome.
Real Enterprise Problems
Every build begins with a real business challenge, researched end-to-end, defining the user journey, and identifying where AI can create meaningful improvements.
Build at the Frontier
Our engineers design and build working programs using the latest AI technologies and partner platforms. Each POC explores how AI can address a specific enterprise use case.
Prove What's Possible
Every POC becomes a story of the problem, the build, and the thinking behind the solution, showcasing what Winter engineers can create with AI.
25 builds. Real partner technologies. Enterprise problems.
- Day 1 — AI-Based Self-Healing with CodeGen by testRigor, built with testRigor (CodeGen (autonomous test fix + PR); testRigor MCP)
- Day 2 — Peer AI MongoDB Migrator, built with MongoDB (MongoDB / Atlas, Atlas Vector Search, MongoDB MCP)
- Day 3 — LangChain Deep Agent POC: Intelligent AML Compliance Workflow, built with LangChain (DeepAgent on LangGraph; LangChain agents; RecursiveTextSplitter)
- Day 4 — Live Flight Concierge: Real-Time AI Assistant for Travelers, built with MongoDB (MongoDB / Atlas, Atlas Vector Search, MongoDB MCP); LangChain (DeepAgent on LangGraph; LangChain agents; RecursiveTextSplitter); Confluent (Confluent Cloud, Flink, Confluent MCP (Kafka))
- Day 5 — CareDocs IQ: Intelligent Healthcare Document Parser, built with MongoDB (MongoDB / Atlas, Atlas Vector Search, MongoDB MCP); LlamaIndex (LlamaParse)
- Day 6 — AI-Powered Terraform Version Migration Impact Analyzer
- Day 7 — Agentic UI Validation Leveraging LLMs and testRigor MCP, built with testRigor (CodeGen (autonomous test fix + PR); testRigor MCP)
- Day 8 — Multimodal RAG for Trusted Knowledge Management
- Day 9 — Graph Native Clinical Risk Reasoning Using Neo4j Aura Agent and GraphRAG, built with Neo4j (Neo4j AuraDB + Aura Agent, GraphRAG)
- Day 10 — Durable AI-Driven Credit Decisioning using Temporal Workflows and Azure OpenAI, built with Temporal Technologies (Temporal durable workflows)
- Day 11 — Real-Time Payment Operations Copilot using Confluent Cloud, Apache Kafka, Confluent MCP, and Azure OpenAI, built with Confluent (Confluent Cloud, Flink, Confluent MCP (Kafka))
- Day 12 — AI Governance Firewall: An Engineering Approach to Secure Enterprise LLM Adoption
- Day 13 — Microsoft NLWeb for an Agent-Ready Publishing Website
- Day 14 — Agent Regression Harness, built with LangChain (DeepAgent on LangGraph; LangChain agents; RecursiveTextSplitter)
The work, day by day.
AI-Based Self-Healing with CodeGen by testRigor
- The Challenge
- QA teams rely on manually written tests and bug fixes that quickly go stale, slowing development cycles and letting bugs slip through to production undetected.
- The Build
- An AI-powered CodeGen feature from testRigor that autonomously detects bugs on a live application, generates fixes, and pushes the corrected code as a pull request to GitHub in real time.
- Business Impact
- Automates the full fix-and-commit cycle, reducing manual debugging effort.
- Speeds up bug resolution from detection to pull request.
- Enables self-healing QA with minimal developer intervention.
- Creates an auditable trail from bug detection through to the fix.