Jeremie Guedj - July 20th 2023

My primary goal is to ramp up on a complete tech stack. I will achieve this goal by solving a problem I personally face, so I won’t be investing time nor thought into scalability, market size etc.

I have listed 4 problems that I face as a PM at Meta, and filled a decision matrix to help me select which to solve.

Each problem listed also has the required tech blocks required to build it, colored in green when available, or red when not available/accessible

[👑 Winner] Problem 1: Creating GOOD framework docs to drive decisions takes too much effort

• Working on projects involving lots of people from multiple orgs and multiple timezones.
• As a result, a lot of the communication and decision making has to be made asynchronously.
• One of the most efficient ways of doing so is to have a common document-based framework to drive the decision process, i.e. prioritization or tradeoff decisions.
• However, creating docs driving processes or decision takes a long time and/or lots of cognitive effort, but its worth it: a well-crafted framework will drive fast progress, quick alignments and clarity to all stakeholders On the other end a bad framework will waste a ton of time to everyone involved, with endless meetings and tons of comments to address etc

| Frequency | Paint Point Magnitude | Impact Estimation if solved | High Level Solution | Tech Blocks Required | | --- | --- | --- | --- | --- | | Daily/Weekly | High - cognitive effort, time investment and resulting opportunity cost | High - given the frequency and paint point magnitude, and the positive effect of a well-crafted framework | Phase 1 (MVP)

• A searchable repository of pre-made framework documents incl. metadata on the framework

Phase 2 - Craft a framework doc by choosing logical blocks i.e. “tradeoff table” that are created with parameters from user input (i.e. the different options of a tradeoff table)

Phase 3 - Add private and public doc lists, allow users to upvote docs created by others

Phase 4

• Allow completely dynamic doc creation based on free text user input and GPT4 API | - User authentication and management

• Google APIs for docs creation

• OpenAI GPT3.5/4 API

• Fullstack basics: front end, front end logic, backend, DB

👍 All tech blocks are immediately accessible |

Problem 2: Mountains of updates to keep up with around the clock

• In a company like Meta, an unimaginable volume of relevant information is being created by peers every day, in the form of Workplace posts (an internal Facebook for companies), posts that will usually have discussions/questions in the comment thread.
• As a PM, there is a need (and expectation) to stay on top of everything relevant to my product, that includes reading mountains of posts and their comments, asking and answering questions etc.
• Since I work in a multi-timezone org, this is around the clock, and I’ll get up in the morning with a fresh pile of posts created overnight in the West Coast.
• Keeping track of all relevant posts is very challenging, and sometimes even impossible. This is very time consuming, and often important discussion can be missed or “discovered” late.