emilià

T2 · manager / operator

CTO of Factorial, a Catalan HRIS/platform company with 16,000+ clients across 9 markets, $100M+ ARR, and 25+ integrated products covering HR, finance, IT, and talent management.

4 calls·4 names·100% bull·last heard 3 months ago·itnig
track record

no scored calls yet — needs a stated position or a categorical verdict, with a matured window vs SPY

top calls

highest conviction · one per company
1sthigh conviction
$DUCKDBDuckDB

Factorial CTO says DuckDB 'worked magic' for analytical queries in AI agent loops

Embedding DuckDB in-process allows AI agents to offload complex analytical SQL (joins, aggregations) from LLM context, solving token overflow and reasoning errors while maintaining dynamic schema discovery.

itnig2026-05
2ndmedium conviction
$COPILOT-KITCopilot Kit

Factorial uses Copilot Kit to bridge AI agent actions with browser UI state

Copilot Kit enables shared state between LLM and browser, allowing agents to trigger frontend mutations (popups, form fills) via WebSocket-streamed tool calls.

itnig2026-05
3rdmedium conviction
$MASTRAMastra

Factorial CTO endorses Mastra JS as sweet-spot framework for building AI agents

Mastra provides high-level primitives without exposing full framework complexity, enabling rapid iteration on agent architecture with responsive team support.

itnig2026-05
4 total
$CLICKHOUSE
ClickHouse
LOWemilià·itnig·3 months ago
Factorial uses ClickHouse for high-cardinality analytics beyond DuckDB capacity
ClickHouse handles extremely high cardinality datasets (millions of rows) that exceed DuckDB's in-process capacity, serving as the analytical backend for queries with efficient GraphQL endpoints.
"we obviously have a ClickHouse for something that's extremely high cardinality that we cannot dump into the DuckDB because it's like millions and millions of rows. So, for this ca…"
40:49
$DUCKDB
DuckDB
HIGHemilià·itnig·3 months ago
Factorial CTO says DuckDB 'worked magic' for analytical queries in AI agent loops
Embedding DuckDB in-process allows AI agents to offload complex analytical SQL (joins, aggregations) from LLM context, solving token overflow and reasoning errors while maintaining dynamic schema discovery.
"And this is kind of a SQLite but for analytics. So, you can attach it into the process. So, you don't need a server, you don't need anything, you can be attached it to the running…"
30:02
$COPILOT-KIT
Copilot Kit
MEDemilià·itnig·3 months ago
Factorial uses Copilot Kit to bridge AI agent actions with browser UI state
Copilot Kit enables shared state between LLM and browser, allowing agents to trigger frontend mutations (popups, form fills) via WebSocket-streamed tool calls.
"We're using Copilot Kit as a UI front end that has a way to share state between LLMs and the browser. And then it's also allows you to call front end tools like inside inside the…"
46:24
$MASTRA
Mastra
MEDemilià·itnig·3 months ago
Factorial CTO endorses Mastra JS as sweet-spot framework for building AI agents
Mastra provides high-level primitives without exposing full framework complexity, enabling rapid iteration on agent architecture with responsive team support.
"we picked Mastra. It's Mastra JS. That's uh it for us it was extremely sweet spot between super low-level things like Versel AI SDK was like 1 year ago because it was just like he…"
13:22