Daniel Avshalom, Parallel Signal — March 2026
MGME has a scientifically validated product with real traction — published in Science, backed by Innosuisse grants, with enterprise clients like EY and hedge funds in the pipeline. The fundamentals are strong.
Two themes came up clearly: the team could get significantly more out of AI tools across their daily workflows, and the marketing engine needs to be built. Both are solvable — and connected. A team that understands AI tools will also be better equipped to scale marketing efforts.
You mentioned this is a make-or-break year. I'd like to help make sure it's a make year. Here's what I'm thinking.
Per-assessment cost in single-digit francs
vs. standard assessment methods
Enterprise partnership potential volume
Lean team, high output required
These came up in our conversation. I'm listing them here so we can confirm which ones matter most to you and the team — everything we build should be based on what you actually need, not my assumptions.
You mentioned these are time-intensive. Could AI help draft prep materials or structure the output faster?
Each enterprise client wants tailored reports. Worth exploring whether AI templating could speed this up significantly.
Currently one person handles certifications. If demand grows, this could become a bottleneck. AI-assisted course delivery might help.
Your data analysts might benefit from AI-powered analysis tools, especially within your existing Microsoft and Google environments.
The team uses Copilot, ChatGPT, and Gemini but there's no shared framework for what's best for what — and what's safe given ISO 27001.
A separate conversation, but clearly important. When the time is right, there's a lot we can do here systematically.
The goal is simple: by the end of this process, each person on your team has hands-on experience with an AI tool that makes their specific work faster or better — and they know how to keep going on their own.
Each team member shares a brief overview: what they do day-to-day, what takes the most time, what they've already tried with AI, and where they think AI could help. This can be a simple form or short 1:1 conversations.
This shapes everything I prepare. No generic content — every demo and example will be based on your team's actual workflows.
Based on what your team shares, I'll build "before vs. after" demos — showing how specific tasks look today vs. how they look with the right AI tools. Real examples, your data patterns, your tools.
Part A: The Landscape. An overview of what's out there — ChatGPT, Claude, Gemini, Copilot, Perplexity — what each one does best, how they name the same features differently, and what's safe for an ISO 27001 company.
Part B: Your Transformations. Live demos of your team's workflows, transformed. "This used to take 4 hours. Watch it happen in 10 minutes."
Part C: Hands-On. Each person implements their own transformation with guidance. Everyone leaves with something working — not just slides.
A summary of what was built, quick wins to implement immediately, and an open channel for questions as the team starts using the tools in their daily work.
A preview of what we'd cover in the workshop. Each tool has strengths — the key is matching the right one to the right problem.
| Microsoft Copilot | ChatGPT | Claude | Google Gemini | Perplexity | |
|---|---|---|---|---|---|
| Best for | Teams on Microsoft 365 | General tasks, coding | Deep analysis, writing, system building | Google ecosystem | Research with citations |
| Custom assistants | Copilot Studio | Custom GPTs | Projects | Gems | — |
| Company context | Reads SharePoint, Outlook, Teams | Custom Instructions + uploads | Projects with persistent memory | Workspace integration | — |
| Data analysis | Excel Agent Mode (strongest) | Code Interpreter | Artifacts + analysis | Sheets AI | Limited |
| Security | ISO 27001, SOC 2, Purview DLP | Enterprise tier only | Enterprise tier available | SOC 1/2/3, ISO 42001 | Basic |
| Data used for training? | No | Enterprise: No | No | No | Check ToS |
| Meetings | Teams (excellent) | — | — | Google Meet | — |
| Automation | Power Automate | Zapier / API | MCP servers / API | Apps Script | API |
| Concept | ChatGPT | Claude | Gemini | Microsoft |
|---|---|---|---|---|
| Custom AI assistant | Custom GPTs | Projects | Gems | Copilot agents |
| System instructions | Custom Instructions | CLAUDE.md | Gem instructions | Studio prompts |
| File analysis | Code Interpreter | Artifacts | Code execution | Excel Agent Mode |
| Company knowledge | GPT Knowledge files | Project docs | NotebookLM | SharePoint + Copilot |
| Automation | Zapier / API | MCP servers | Apps Script | Power Automate |
This would be customized to MGME's specific tools and security requirements during the workshop.
To move this forward:
Both Microsoft and AWS have startup programs that could give you significant free credits. Microsoft for Startups offers up to $150K in Azure credits and 50 free M365 Business Premium seats. AWS Activate offers up to $100K in credits.
You mentioned you use both platforms. Worth checking eligibility for both — the application takes about 15 minutes.
Let me know if this captures our conversation correctly.
Happy to adjust the approach based on your feedback.