A passion for finding patterns and new perspectives

TL;DR: Giv­ing LLMs access to a knowl­edge base is easy. Giv­ing them scoped, per­mis­sioned, read/write access with­out hand­ing over the keys to the king­dom is hard. Vault Bridge is a Rust mid­dle­ware lay­er that sits between an Obsid­i­an LiveSync data­base (CouchDB) and an MCP client, enforc­ing pol­i­­cy-as-code bound­aries. The code is available…

zkPass allows to unlock web2 data for web3 appli­ca­tions, with pri­va­cy. Scroll down to get the full tech­ni­cal deep dive. What is zkPass? zkPass is a pro­to­col that uses Zero-Knowl­­edge proofs to let you prove facts about your pri­vate Web2 data (from any HTTPS web­site) to Web3 appli­ca­tions. You can…

DE is hard and often requires for ad-hoc code changes to deal with the idio­syn­crasies of the spe­cif­ic dataset. LLMs could offer a big help in assist­ing and even­tu­al­ly automat­ing these tasks. The big play­ers to watch out for are those with estab­lished low-code DE solu­tions, espe­cial­ly those already intimate…

I’ve heard some valid crit­i­cism of blockchain tech: it is a very inef­fi­cient data­base, there’s noth­ing sub­stan­tial­ly new peo­ple are doing with it, DeFi is self-ref­er­en­­tial. All of these have some truth in them. On top of this the indus­try is over-hyped and inun­dat­ed by star­tups with mil­lions in funding…

Top­ic Mod­el­ing is an unsu­per­vised learn­ing pro­ce­dure sim­i­lar to clus­ter­ing. In this post I give a brief overview of Top­ic Mod­el­ing, present my library trans­former­topic and apply it to scraped news about COVID

Top­ic Mod­el­ing is an unsu­per­vised learn­ing pro­ce­dure sim­i­lar to clus­ter­ing. In this post I give a brief overview of Top­ic Mod­el­ing, present my library trans­former­topic and apply it to scraped news about COVID

I was plan­ning to add pro­file pic­tures to my now defunct social-app for musi­cians Riff on That and I want­ed to offer a ran­dom­ly gen­er­at­ed pic­ture as a default option. I had a look around for exist­ing solu­tions in python and could­n’t find any­thing to my lik­ing, so I came…

Two learn­ing par­a­digms The tra­di­tion­al par­a­digm for learn­ing is how all Schools and Uni­ver­si­ties work: at the out­set you define a step by step pro­gram and you fol­low it lin­ear­ly. Each chunk of the learn­ing builds on the pre­vi­ous and you are guar­an­teed that, if you fol­low it dili­gent­ly until…