Skip to content
Last run passedcali
UTC--:--:--

← startups

Mnemonic Labs

no website yet

via ycombinator.com · Nov 15
54/100YC

A foundation model for quantitative trading that learns directly from market activity

  • Python
  • PyTorch
  • TensorFlow
  • CUDA
  • AWS

bottleneck › scaling the data ingestion and preprocessing pipeline for massive market datasets

pitch

You could replace the custom data pipeline with a Next.js front‑end that streams market data into a vector store, letting you query and enrich raw ticks on the fly. This lets your engineers focus on model research while the semantic search layer handles fast, scalable retrieval of relevant market contexts.

email draft › no published address

intent: build for them (a concrete bottleneck on a stack you can work with)

Mnemonic Labs: a thought on scaling the data ingestion and preprocessing pipeline for ma

Hi Videh,

Congrats on the launch (https://www.ycombinator.com/launches/XRK-mnemonic-labs-openai-for-financial-markets).

I came across Mnemonic Labs: A foundation model for quantitative trading that learns directly from market activity.
One thing I noticed: scaling the data ingestion and preprocessing pipeline for massive market datasets.

You could replace the custom data pipeline with a Next.js front‑end that streams market data into a vector store, letting you query and enrich raw ticks on the fly.

Would a short trial project to tackle this be useful?

If this isn't relevant, just reply "no" and I won't follow up.

signals

qualification

  • fit10000%
  • reach0%
  • evidence5300%

timeline

  1. Nov 15 found · Stage: YC · Product: A foundation model for quantitative trading that learns directly from market activity · Bottleneck: scaling the data ingestion and preprocessing pipeline for massive market datasets · +4 moresource
  2. Nov 15 learned · Previously at: Kalki Seksaria previously_at Tower Researchsource
  3. Nov 15 launch · Launchsource

similar companies

score › fit 25/25 · identity 9/15 · people 10/20 · connections 10/20 · contact 0/20

missing › location · funding amount · people profiles · investors · email · phone

people

  • Videh Seksaria · Co‑founder
    “We’re Mnemonic Labs - a neolab training a foundation model for quant trading that learns directly from market activity”

connections

  • Videb Seksaria founded Mnemonic Labs
    “We’re Mnemonic Labs - a neolab training a foundation model for quant trading”
  • Kalki Seksaria cofounded Mnemonic Labs
    “We met 28 years ago shortly after Videh was born (we’re brothers!)”

Trackbacks

Company

  • core_product: A foundation model for quantitative trading that learns directly from market activity inferred · first seen Nov 15, last Nov 15
  • funding_stage: YC inferred · first seen Nov 15, last Nov 15
  • likely_bottleneck: scaling the data ingestion and preprocessing pipeline for massive market datasets inferred · first seen Nov 15, last Nov 15

People

  • Videh Seksaria · Co‑founder verified · first seen Nov 15, last Nov 15
    “We’re Mnemonic Labs - a neolab training a foundation model for quant trading that learns directly from market activity”

Connections

  • Kalki Seksaria cofounded Mnemonic Labs verified · first seen Nov 15, last Nov 15
    “We met 28 years ago shortly after Videh was born (we’re brothers!)”
  • Videb Seksaria founded Mnemonic Labs verified · first seen Nov 15, last Nov 15
    “We’re Mnemonic Labs - a neolab training a foundation model for quant trading”
  • Videh Seksaria previously at Citadel verified · first seen Nov 15, last Nov 15
    “Videh was the youngest quant at Citadel at 19”
  • Kalki Seksaria previously at Tower Research verified · first seen Nov 15
    “Kalki made $20M in annual PnL at 22 at Tower Research”