Running an AI Music Project: Week 5 Progress Report
Clafacio Lobo · Project Manager, Musinique · Humanitarians.ai
Five weeks in. This is the week I want to talk about what it means to run a project when the pieces are all moving but not all of them are landing yet and what the gap between output and completion actually looks like from the inside.
The Nana Series — From Concept to Asset
Last week I wrote about the Nana educational content series as a coordination milestone three contributors working toward a single shared deliverable for the first time. This week that work moved from coordination into execution, and the results are starting to come together.
Nidhi produced an original educational track on Suno a song built for children aged 3 to 5, using the Ghost Artist Nana as the creative identity. The track is live. Lyrics are original. The YouTube video is in progress animations are ready, the song is ready, and the edit is the remaining step.
Ragamalika is creating images on Midjourney to support the children’s content direction, developing the visual identity that will accompany the Nana series going forward. That work is in progress and the direction is clear.
Sakshi reviewed the animations for age-appropriateness a specific, necessary quality check before the video goes public. She also followed up on the annotator shortlist for the playlist scoring validation, picking up the thread from Shruti’s earlier outreach.
What this week showed is that the Nana series is real. It is not a concept anymore. There is a track, there are visuals in development, and there is a video close to completion. The question for next week is whether all three pieces come together into a single publishable asset and whether the YouTube upload actually goes live.
The Research Paper Is Waiting On One Thing
The research pipeline is in a holding pattern that is not anyone’s fault and is not a failure but it is worth being honest about.
Nixon built the results section template. Shruti filtered the literature review gaps and drafted a pitch that is ready to submit. The methodology is defined. The scoring system is designed. The annotator outreach is active. Everything that can be done without Artist.tools data has been done.
The paper is waiting on that data. Until it arrives, the results section stays empty and the submission timeline stays uncertain. That dependency has been on the risk board since Week 2. It is still there.
What changed this week is that the team stopped waiting passively and started building the infrastructure around the gap templates, annotations, validation frameworks so that when the data arrives, the turnaround to a submittable draft is as short as possible. That is the right response to an external dependency you cannot control.
The annotator search is the newest active risk. Finding people willing and qualified to validate a playlist scoring system takes time, and the timeline for the paper depends on it. Sakshi is following up on the shortlist. I will have a clearer status on this by end of next week.
Publishing — Volume Is Strong, Distribution Needs Attention
The content output across the team remains high. Articles are publishing on Musinique.net and the Humanitarians.ai Substack consistently. Nixon’s timing experiment publishing on weekends to test for better click-through rates is underway, and I am watching for early signals on whether it is working.
But I want to be honest about something. High output volume is not the same as growing reach. The question I am asking more seriously now at Week 5 is whether the content we are producing is finding its audience or just accumulating on the site. That is a distribution question, not a production question, and it is one the team has not fully turned its attention to yet.
The upload and documentation process is still the place the project leaks operationally. I have named this gap before. This week I tracked confirmations more directly than in previous weeks and the picture is cleaner but not clean. It is still the thing I spend more coordination time on than I should.
What Week 6 Needs to Deliver
Three things matter most going into next week.
The Nana YouTube video needs to go live. The pieces are all built track, animations, visuals. The edit and upload are the remaining steps. Done is a published YouTube link confirmed and shared with the team.
The annotator outreach needs a status with a number. Not “in progress” how many people have been contacted, how many have responded, and what is the realistic timeline for having a validated annotation set. That is the information the paper timeline depends on.
And the Artist.tools pitch needs to go out. It has been ready to submit for two weeks. The data dependency is real but it should not be holding up the pitch itself. Sending the pitch and receiving the data are not the same action. That pitch goes out next week.
Five weeks in, this project is producing real work. The gaps are smaller than they were at Week 2. The coordination is tighter. The creative direction is clearer. But there is a difference between a project that is running and a project that is landing, and Week 6 is where I want to close that gap.
More next week.
Clafacio Lobo is the Project Manager, Musinique Humanitarians.ai. Follow the project at musinique.net · humanitarians.ai/clafacio-lobo