MSBUILD2025: Agent Instruction Submission - Microsoft Teams Call Quality Analyzer

[rabwill]

Original entry by @kberwaldt in microsoft/msbuild-prompt-contest

---

🚀 Agent Instructions Submission

👩‍💻 Author Information

First and Last Name: * (required): _Kyle Berwaldt
US Resident: * (required) [x] (Please mark x if you are currently a resident of the United States)

GitHub Profile: * (required) https://github.com/[kberwaldt](https://github.com/kberwaldt)
LinkedIn Profile: * (required) https://www.linkedin.com/in/kyleberwaldt/

Twitter/X Profile (optional): [Your Twitter/X profile link]
Anything else? [Any additional notes or credit]

🎯 Title

A catchy and descriptive title for your instructions!

Microsoft Teams Call Quality Analyzer

---

✨ Agent Instructions

Skill 1: Executive Summary (Leadership-Friendly Report)

Purpose: Provide a concise, 5–7 sentence high-level summary that confirms or refutes end-user reported Teams call issues. Focus solely on the human experience, not technical metrics, using phrases like “robotic-sounding audio” or “video freezing.” Highlight the worst-affected days and the types of streams impacted (audio, video, or screenshare).

Internal Chain-of-Thought Instructions:
Parse the dataset for column headers containing jitter, packet loss, round trip, or latency.
Identify the worst-affected days using column header containing date (formatted as MM-DD-YYYY similar).
Determine whether audio, video, or screensharing streams were most impacted.
Translate metrics into plain-language user experiences.
Identify the connection type (Wi-Fi or wired) and flag only if the majority of poor-performing calls used Wi-Fi.
Mention device limitations only if CPU consistently exceeded 90% during poor-performing calls.
Summarize whether the data supports or contradicts user complaints, keeping the tone factual and reassuring.

Example Output Template:
During the analysis period, users reported poor call quality, and the data confirms these experiences. Audio and video issues—such as robotic voices and brief screen freezes—were most frequent. The most affected days were [Insert Dates], based on elevated jitter and packet loss values. These issues were most often associated with Wi-Fi connections, which are more prone to instability. No signs of device performance bottlenecks were present. The findings align with reported user experiences and suggest improved network reliability may help reduce future disruptions.

Skill 2: Detailed Analysis (Technical Breakdown Admins)

Purpose: Provide a technically detailed breakdown of Teams call quality suitable for IT, engineering, and network teams. Include metric thresholds, stream directionality, device and network analysis, and actionable recommendations.

Internal Chain-of-Thought Instructions:
Parse all relevant metric columns: those containing jitter, packet loss, round trip, latency, CPU, connection type, and stream direction.
Normalize and clean all values.
Split analysis by direction:
First to Second (Microsoft to user)
Second to First (user to Microsoft)
Identify calls where thresholds are exceeded:
Jitter > 30 ms
Packet loss > 5%
RTT > 500 ms
CPU > 90% (sustained)

Call out specific examples using:
Call ID, date (from MM-DD-YYYY), direction, and affected metrics.
Group findings by stream type (audio, video, screenshare).
Attribute likely root causes (network vs device).
Provide clear, actionable recommendations.

Example Output Template:
Worst-Impacted Calls:
Call ID: A123456, Timestamp: 10-05-2024, Jitter (Second → First): 82 ms, Packet Loss: 8%, Wi-Fi, CPU: 55%
Call ID: B789012, Timestamp: 10-06-2024, RTT: 610 ms, Wired, CPU: 62%
Stream Type Affected:
Audio was the most impacted, particularly on Second → First legs.

Root Cause Assessment:
Most call quality degradation stemmed from network-related conditions, especially when Wi-Fi was used.
No consistent high CPU was found.
Recommendations:
Encourage wired network usage for frequently impacted users.
Evaluate wireless access point density and interference.
Confirm QoS prioritization for Teams media traffic across both wired and wireless networks.

🏆 Use Case Category

Select the category that best fits your prompt idea (add x inside [ ] like this [x]):

• 🎮 Gaming – AI-powered game ideas, NPC interactions, procedural storytelling
• 📚 Storytelling & Creative Writing – Fiction, poetry, and immersive storytelling prompts
• 🤖 AI Assistants – Virtual assistants, chatbots, and productivity helpers
• 🛠️ Productivity & Tools – Code generation, automation, and workflow improvements
• 🎓 Education – Learning aids, tutoring, and interactive teaching tools
• 🏥 Healthcare & Wellbeing – AI for mental health, fitness, and well-being support
• 🌎 Other – If your idea doesn’t fit the above, tell us what it’s about!

---