# Background On March '23 I (@mpj) reached out to Daniel on my interest to build a startup on the AI-space. I am confident because I'm equipped with - My little fundraising experience helping startups raising in Stacks Accelerator in 2021. - A spark of energy and excitement around AI when ChatGPT hyped. - The idea is to figure out early traction and then fundraise. Daniel offered me a funding of $25k for two projects - which involved AI. This offer is so good and I took it. # What was the idea Original idea from me: - AI girlfriend / AI friend - Tools to help with AI art (prompt playground) Idea from Daniel & Nicole: - NFT Marketplace on Gnosis (taken but dropped) - GnosisGPT (not taken) - CrewAI (taken but pivoted) - Hologram (run until July '23) # Funding till-date We spent $11666.9 funding from Daniel and Nicole and another self-funding of $10k from our own pocket totalling $21666.9 in funding. # The longest running experiment: Hologram AI The idea of Hologram is to provide user with ways to generate image easily using pay-to-generate with crypto and easy NFT minting. The idea is people are interested in minting AI-generated image if the image is beautiful or have some art value. # What we don't accomplished - We don't experiment fast enough on Gnosis GPT and NFT Marketplace for Gnosis - Deliver CrewAI as fast possible when it was still hyped - We then pivoted to Hologram but couldn't delivery web3-related feature on time since most of our team member is new to Web3 and smart contract. While I don't have enough time to be very hands on while doing research and finding business opportunities. # What we have accomplished - Won the Gateway.fm challenge in the [Gnosis x Encode Hackathon](https://medium.com/encode-club/encode-x-gnosis-hackathon-prizewinners-and-summary-c8e834aea325), winning $1k of bounty. - Acquired paying user, in-total only [$2 in revenue](https://gnosisscan.io/address/0xd54E6A61332657eCac42146f226e44C6166C86bE) - Getting some community members in [Discord](https://discord.com/invite/Nf6Pb7sfuJ) around 30+ users. - Launched our waitlist program and [alpha application](https://tryhologram.art/) - In our database in total we have 81 users and 18 users in NFT waitlist. - Have people [minted our NFT](https://niftyfair.io/gc/collection/0x45632949a5b0e576914232b2908f84d5c5ef005c/). About 10 NFTs are minted by external users. # What we should do differently - Having complete cross-functional team in the beginning. The lack of good designer is one of the factor we can't delivery good product faster. - Startup is time consuming so I should do it with more time commitment. Day job + new born + startup nights and weekends is something I couldn't do long-term. - Pick the right team. The team that I have when running Hologram, while very nimble, is not setup for success. For example: - I should find better partner - Should have engineer with web3 background - Should have more senior engineer that can delivery with more autonomy # What I will do next - Take a break - Researching on new things that fit my experience and strength to work on. - Build then fundraise, not the other way around # Summary Doing startup is hard, failure is expected and the best way to handle failure is to stop the bleeding before it's too late. In this post-mortem I shared how Hologram failed to deliver the goal of buildling an AI tool that user wants either for conversational bot or for Art (Hologram). While I learned more about what can make Hologram a good product, I realized that Hologram team as of today is not positioned for success and not very comptitive. Thus, before the bleeding not cureable, I decided to pull the trigger and go back to the drawing board. I will be taking a break, researching what I will do next and most definitely building a new stuff that I hope people want. I forever grateful for Daniel and Nicole who give us chance to try to innovate and also to the team who have faith in me to make it work. I proud to what we have accomplishd so far and very inspired by the team dedication on delivering product at their limit of capability.