Arkash Jain.
AI researcher (Harvard) and forward-deployed engineer at Benmore — three published papers including SpatialDINO (3D self-supervised vision transformer for live-cell microscopy), and the toolkit that makes SMB AI consulting compound.
The Complete AI Hardware Stack — Layer by Layer
Apr 2026 · Knowledge
Head of FDE, Benmore
Apr 2026 · Work
$150k total -> $150k every 15 days
2025-2026 · Milestone
Weekly updates
rolling log.
Hiring kickoff, Latinovation, Legora culture, and mech-interp via Kalman + PCA
PCS PR queue cleared across five threads (Trevin, Brad & Eddie, Propurti, PCS checks, ongoing dev), Chantal intelligence-tab handed off and Monica research kicked off, Alex Titov starting Monday, hiring conversations opening with Martins via Trueup + Builtin, and a reading week on Legora's lean Stockholm culture, mechanistic interpretability via Kalman filters + PCA, the Karg port endgame, and Nvidia carving 20% of die area for Groq while positioning Vera CPU for agentic + RL workloads.
Read the log →
PCS hardening week, Anthropic's J-curve, ERCOT vs PJM, and the silicon shortage
PCS audit logging + Kong gateway + JWT hardening, CRL/Workforce shipped, four new engineers onboarded (Jacob, Leo, Aditya, Enrin), and a deep reading week on the AI silicon shortage, Anthropic's 12× ARR jump, ERCOT vs PJM capacity markets, and the SEC/CFTC split on tokenized securities.
Read the log →
Live activity
Refreshed straight from the source.
Public commit cadence plus a private snapshot from a personal commit-graph tool.
Contribution graph · last 12 months
Snapshot · ~April 14, 2026
- 30d commits624
- 7d commits186
- Active days28
- Raw lines823,824
- Adj output520,582
Aggregated across public + private repos.
Commit-tracker · 60d

Physics → VC → Distributed Systems → Harvard AI → Building
I came to Boston from India in 2020. NSF UROP scholar in chemical physics, two stints at Battery Ventures, distributed-systems research at BU, then Harvard's Kirchhausen Lab to build SpatialDINO. Now Head of FDE at Benmore.
Papers
3
Classes Deep-Dived
9
Repos
25+
3 published papers
SpatialDINO: Self-Supervised Learning for 3D Vision Transformers
Arkash Jain et al. (Kirchhausen Lab, Harvard Medical School)
A 3D self-supervised vision transformer for label-free segmentation and tracking of subcellular dynamics in lattice light-sheet microscopy (LLSM). Beat a prior approach co-led by Nobel laureate Eric Betzig on downstream subcellular structure prediction.
Read paper →Close-up of Vesicular ER Exit Sites by Volume Electron Imaging using FIB-SEM
Kirchhausen Lab
Volumetric reconstruction of mammalian ER exit sites at unprecedented resolution via FIB-SEM and learned segmentation.
Read paper →Ultrafast 2DIR comparison of rotational energy transfer, isolated binary collision breakdown, and near critical fluctuations in Xe and SF6 solutions
Arkash Jain et al.
First-author work on supercritical-fluid dynamics using ultrafast two-dimensional infrared spectroscopy.
Read paper →