M37Labs Logo

M37Labs Feed

Knowledge base

Direct from our engineering and research Slack channels. Raw insights, technical breakdowns, field notes, and real-time updates from the M37Labs team.

M

Team M37Labs

knowledge-base
9 months ago

Lot of hype around - claude code + opus 4.5 integrated performance.

https://www.interconnects.ai/p/claude-code-hits-different

Lot of hype around - claude code + opus 4.5 integrated performance.
and it's been said by experts - "Software is becoming free and human design, specification, and entrepreneurship is the only limiting factor."
interconnects.ai

Claude Code Hits Different

Coding agents cross a meaningful threshold with Opus 4.5.

www.interconnects.ai

Preview image for Claude Code Hits Different
Link
Feb 2, 2026 • 6:59 AM
M

Team M37Labs

knowledge-base
9 months ago

AMP - The Agentic Consulting Landscape (Q1 2026) | Jimmy Bijlani | 422 comments

https://www.linkedin.com/posts/jbijlani_amp-the-agentic-consulting-landscape-q1-ugcPost-7422317005527044096-Xz7q?utm_source=share&utm_medium=member_ios&rcm=ACoAAAAaJJEBNJROdSFmWPtf7nsb9a6GVO2GKqo
linkedin.com

AMP - The Agentic Consulting Landscape (Q1 2026) | Jimmy Bijlani | 422 comments

After surveying 250+ consulting and professional services firms, we’ve built one of the clearest views available on how AI is *actually* being adopted - and the gap is STRIKING. Not because firms aren’t using AI, but because the distance between using AI and actually changing how work gets delivered is FAR wider - and firms that delay will find it increasingly hard to catch up. At AI Momentum Partners (AMP), our team pulled together a data-backed landscape of AI adoption and execution in consulting, informed by: - A recent survey of 250+ consulting and professional services leaders (primarily mid-market, but still relevant to anyone from solopreneurs to Big 4) - Direct work with ProServe firms & teams ranging from ~50 to 5,000 people - What we’re seeing inside live delivery environments while building agents for clients We’re sharing the full report for FREE - because we truly believe that 2026 is the inflection point. Here were the 4 most surprising insights: :one: AI adoption is high; transformation readiness is not. Nearly 4 out of 5 consulting firms are already experimenting with AI or running pilots. But fewer than 5% have redesigned their operating models around agentic ownership, supervision, and delivery governance. In 2026, the move from pilots to production will separate firms that operationalize AI from those that fall behind - and the gap between them will widen faster than most expect. :two: Agents are already creating value, but the impact shows up unevenly. Where agents are embedded into real delivery workflows, firms are seeing substantial reductions in production time and early margin impact (typically in the low double-digit range). A small but growing set of professional services firms are also experimenting with packaging agents as “billable team members” - an early signal of how staffing and pricing models may evolve. :three: The agent landscape is emerging in waves, not all at once. We’re seeing a broad set of consulting-specific agents emerge. Some categories (i.e. research, drafting, QA, etc.) are already maturing today, while others are still early and likely to take shape over the next 12–24 months. The firms pulling ahead are sequencing this evolution deliberately instead of treating all agents as interchangeable. :four: Agents are becoming part of client offerings, not just internal tools. Leading firms are embedding agents directly into client-facing work - enabling diagnostics, due diligence, and benchmarking to be delivered in a fraction of the time and cost traditionally required (sometimes 60–80% faster). This is where agents begin to reshape what firms sell, not just how they operate. If you want the full report: like this post, send me a connection request, and comment "consulting" below. And if you repost, I'll move you to the front of the line. These results shocked me a little bit - I’d love to get your take on what surprised you (or rather, didn't surprise you). :point_down: | 422 comments on LinkedIn

www.linkedin.com

Preview image for AMP - The Agentic Consulting Landscape (Q1 2026) | Jimmy Bijlani | 422 comments
Link
Jan 30, 2026 • 3:32 AM
M

Team M37Labs

knowledge-base
10 months ago

Beyond OCR: Agentic Document Extraction with David Park & Andrea Kropp | DeepLearning.AI…

https://www.linkedin.com/posts/ocr-optical-character-recognition-can-read-ugcPost-7417242222531657728-Xjgz?utm_source=social_share_send&utm_medium=ios_app&rcm=ACoAAAAaJJEBNJROdSFmWPtf7nsb9a6GVO2GKqo&utm_campaign=copy_link
linkedin.com

Beyond OCR: Agentic Document Extraction with David Park & Andrea Kropp | DeepLearning.AI posted on the topic | LinkedIn

OCR (optical character recognition) can read text from documents, but it often misses what matters most: layout and visual structure, like tables, charts, forms, and reading order. In "Document AI: From OCR to Agentic Doc Extraction," a new short course built with LandingAI, you’ll learn how to go beyond OCR using Agentic Document Extraction (ADE): parsing documents as visual objects and extracting structured Markdown and JSON, grounded to specific regions on the page. Taught by David Park and Andrea Kropp. Start now: <https://hubs.la/Q03-LSHX0> | 27 comments on LinkedIn

www.linkedin.com

Preview image for Beyond OCR: Agentic Document Extraction with David Park &amp; Andrea Kropp | DeepLearning.AI posted on the topic | LinkedIn
Link
Jan 19, 2026 • 3:22 AM
M

Team M37Labs

knowledge-base
10 months ago

Recursive Language Models: the paradigm of 2026

Recursive Language Models: the paradigm of 2026 https://share.google/kiHrdMPHbm7RfiUp6
primeintellect.ai

Recursive Language Models: the paradigm of 2026

A blueprint for “context folding”: recursively compressing and reshaping an agent’s own context to prevent context rot and keep ultra-long, multi-step rollouts cheap and reliable without relying solely on external file-based scaffolding.

share.google

Preview image for Recursive Language Models: the paradigm of 2026
Link
Jan 17, 2026 • 5:02 PM
M

Team M37Labs

knowledge-base
10 months ago

Adapting the Facebook Reels RecSys AI Model Based on User Feedback

https://engineering.fb.com/2026/01/14/ml-applications/adapting-the-facebook-reels-recsys-ai-model-based-on-user-feedback/
Engineering at Meta

Adapting the Facebook Reels RecSys AI Model Based on User Feedback

We’ve improved personalized video recommendations on Facebook Reels by moving beyond metrics such as likes and watch time and directly leveraging user feedback.  Our new User True Interest Survey (…

engineering.fb.com

Preview image for Adapting the Facebook Reels RecSys AI Model Based on User Feedback
ThreadLink
Jan 17, 2026 • 9:55 AM
M

Team M37Labs

knowledge-base
10 months ago

Reliance halts cell-making plans after failed bid for China tech

https://economictimes.indiatimes.com/industry/cons-products/electronics/reliance-halts-cell[…]er-failed-bid-for-china-tech/articleshow/126473536.cms
The Economic Times

Reliance halts cell-making plans after failed bid for China tech

Reliance Industries has paused its lithium-ion battery cell manufacturing plans in India after failing to secure Chinese technology, highlighting supply chain challenges. The company is now focusing on assembling battery energy storage systems, a move mirrored by other Indian conglomerates facing similar technology hurdles.

economictimes.indiatimes.com

Preview image for Reliance halts cell-making plans after failed bid for China tech
Link
Jan 12, 2026 • 10:35 AM
M

Team M37Labs

knowledge-base
10 months ago

A multimodal sleep foundation model for disease prediction

https://www.nature.com/articles/s41591-025-04133-4?utm_source=alphasignal&utm_campaign=2026-01-07&lid=8TmBFtVxNWdxwfXe
Nature

A multimodal sleep foundation model for disease prediction

Nature Medicine - A deep learning-based model, developed using the rich, multimodal data available from polysomnography-derived sleep recordings, performs well on common sleep analysis tasks and...

www.nature.com

Preview image for A multimodal sleep foundation model for disease prediction
Link
Jan 8, 2026 • 5:20 PM