PUBLISHED AI RESEARCH AND PROJECTS

AI research and projects

The research, projects, and explanations behind the people. Read the source, inspect verified code and models, and find a collaborator.

A source check confirms that a source supports the summary. It is not an independent replication or endorsement. Our methodology ↗

Showing 1–16 of 567 proofs of work · Newest first
Page 1 of 36
implementationChecked Sep 25, 2026

Accenture Foundation Model Adoption & Value Realization Framework

Lan Guan · Accenture

Developed Accenture's proprietary enterprise AI reinvention methodology, deploying foundation models across supply chain, customer service, and software engineering for Fortune 500 clients while benchmarking return on AI investment (ROAI).

Scope & limitations

Large-scale enterprise legacy IT modernization timelines often outlast fast-moving generative foundation model upgrade cycles.

Context: Enterprise LLM operationalization, multi-cloud foundation model architectures, ROAI metrics.

View mission
explanationChecked Sep 25, 2026

AI Market Structure, Commoditization, and Value Accrual Analysis

Elad Gil · Independent / Venture & Operating Executive

Authored definitive analyses evaluating where economic value accrues across the AI value chain—comparing hardware hyperscalers, foundation model providers, middleware orchestrators, and vertical application incumbents.

Scope & limitations

Qualitative economic framework that depends on rapid infrastructure capex cycles and open-source model capability parity.

Context: Microeconomic modeling of foundation model inference, enterprise gross margins, proprietary data moats.

View mission
explanationChecked Sep 25, 2026

CUDA Accelerated Computing Platform & The Genesis of Generative AI

Jensen Huang · NVIDIA

Pioneered and championed the parallel computing architecture CUDA, articulating how single-instruction multiple-thread (SIMT) hardware execution unlocks orders-of-magnitude algorithmic speedups for deep backpropagation and foundation model matrix multiplication.

Scope & limitations

CUDA vendor lock-in creates significant architectural migration friction for alternate hardware accelerators (TPUs, NPUs) requiring cross-compilation layers like Triton.

Context: CUDA C/C++, cuBLAS, cuDNN, parallel thread execution (PTX).

View mission
explanationChecked Sep 25, 2026

Data-Centric AI and Programmatic Curation for Foundation Model Post-Training

Alex Ratner · Snorkel AI / University of Washington

Authored technical papers establishing the primacy of data curation over model architecture tuning in enterprise LLM adaptation, demonstrating that programmatic alignment yields higher benchmark gains than parameter scaling.

Scope & limitations

Highly specialized enterprise domains (such as clinical medical ontology or complex derivative trading) require expert human-written labeling functions.

Context: Foundation model post-training, DPO/RLHF preference data synthesis, enterprise fine-tuning benchmarks.

View mission
implementationChecked Sep 25, 2026

Databricks Mosaic AI & Lakehouse Architecture for Foundation Models

Ali Ghodsi · Databricks

Led the engineering integration of MosaicML training and fine-tuning pipelines directly into the Databricks Lakehouse, enabling organizations to pretrain and align custom open-weight models securely on internal enterprise data lakes.

Scope & limitations

On-premise and private cloud cluster provisioning constraints can bottleneck distributed tensor parallelism across non-homogeneous GPU topologies.

Context: DBRX, MPT-7B/30B, MosaicML Composer, Delta Lake, Vector Search.

View mission
implementationChecked Sep 25, 2026

Deterministic Agentic Execution Architecture for Enterprise Customer Systems

Bret Taylor · Sierra / OpenAI

Architected Sierra's Agent OS, combining generative LLM natural-language comprehension with deterministic state machines and strict transactional API sandboxes to prevent agent hallucination in consumer-facing enterprise tasks.

Scope & limitations

Enforcing strict deterministic guardrails constrains open-ended agent creativity and requires substantial domain-specific schema modeling.

Context: Multi-LLM routing, deterministic state automata, transactional CRM/ERP webhooks.

View mission
explanationChecked Sep 25, 2026

Early-Stage Scaling and Organizational Architecture for AI Companies

Elad Gil · Independent / Venture & Operating Executive

Published the operational playbook for managing rapid organizational scaling, engineering hiring, and capital allocation during exponential platform transitions.

Scope & limitations

Generic operational guidelines require substantial adaptation for capital-intensive, GPU-constrained AI pretraining labs.

Context: High-growth tech operating principles, governance structures, executive board design.

View mission
implementationChecked Sep 25, 2026

Enterprise Knowledge Graph Grounding for Multi-Source RAG Systems

Arvind Jain · Glean

Architected Glean's enterprise intelligence engine, coupling deep permissions-aware crawling of 100+ SaaS applications with dense vector embeddings and a dynamic enterprise knowledge graph for sub-second grounded generative question answering.

Scope & limitations

Maintaining real-time permission sync across complex Active Directory, Okta, and SaaS permission structures introduces significant distributed indexing overhead.

Context: Hybrid sparse-dense retrieval, enterprise permissions ACL enforcement, LLM synthesis.

View mission
explanationChecked Sep 25, 2026

Generative Adversarial Networks and Deep Generative Modeling Curriculum

Sharon Zhou · Lamini / ex-Stanford

Authored and taught the comprehensive Stanford/DeepLearning.AI generative modeling curriculum, breaking down minimax optimization, Wasserstein loss, and conditional synthesis for over 100,000 developers worldwide.

Scope & limitations

Classical GAN training suffers from mode collapse and training instability compared to contemporary score-based diffusion models.

Context: GANs, conditional image synthesis, Wasserstein divergence, minimax game theory.

View mission
implementationChecked Sep 25, 2026

Lamini Memory Tuning: Precise Fact Recall Without Hallucination

Sharon Zhou · Lamini / ex-Stanford

Developed Lamini Memory Tuning, a technique embedding millions of precise factual enterprise associations into LLM weights with zero hallucination while preserving generalized instruction-following capability.

Scope & limitations

Aggressive factual weight baking risks model overfitting and can degrade out-of-distribution reasoning on non-memorized prompts.

Context: LoRA/PEFT adaptation, cross-entropy weight tuning, hallucination evaluation harnesses.

View mission
explanationChecked Sep 25, 2026

No Priors: Rigorous Technical Dissection of Frontier AI Architectures

Sarah Guo · Conviction / No Priors

Created and hosts the premier AI practitioner podcast, conducting technical interviews dissecting pretraining cluster topologies, inference scaling laws, and enterprise agent adoption dynamics with frontier researchers and CEOs.

Scope & limitations

Podcast interviews capture point-in-time perspectives in an ecosystem where state-of-the-art benchmarks shift on a monthly cadence.

Context: Frontier LLMs, multimodal models, physical AI robotics, synthetic data generation.

View mission
implementationChecked Sep 25, 2026

NVIDIA Blackwell Platform: Accelerated Architecture for Trillion-Parameter Generative AI

Jensen Huang · NVIDIA

Architected the Blackwell B200 GPU and NVLink 5 switch chip infrastructure, introducing second-generation Transformer Engine FP4/FP8 micro-tensor scaling and high-density liquid-cooled rack computing designed to reduce trillion-parameter LLM inference costs and energy consumption by up to 25x.

Scope & limitations

High thermal design power (up to 120kW per NVL72 rack) requires specialized data center liquid-cooling retrofits and high-density power distribution.

Context: NVIDIA Blackwell B200, NVLink 5 (1.8TB/s bidirectional interconnect), FP4/FP8 Transformer Engine, TensorRT-LLM.

View mission
implementationChecked Sep 25, 2026

OpenAI o1 & Test-Time Compute Scaling Architecture

Sam Altman · OpenAI

Directed the commercialization and architectural rollout of OpenAI o1, shifting foundation model capability scaling from purely pretraining tokens to inference-time chain-of-thought reinforcement learning reasoning loops.

Scope & limitations

Increased inference latency (tens of seconds per complex query) and higher token costs make pure test-time reasoning impractical for real-time conversational streaming without hybrid routing.

Context: OpenAI o1, chain-of-thought reasoning, reinforcement learning from human and rule-based feedback (RLHF/RLAIF).

View mission
explanationChecked Sep 25, 2026

Planning for AGI and Beyond: Economic & Societal Governance Blueprint

Sam Altman · OpenAI

Authored OpenAI's strategic roadmap detailing gradual iterative deployment of frontier AI systems, safety red-teaming protocols, empirical alignment benchmarks, and institutional readiness structures necessary for advanced agentic economies.

Scope & limitations

Qualitative strategic governance framework subject to competitive market pressures and rapid geopolitical shifts.

Context: Foundation model safety evals, iterative public deployment, economic displacement modeling.

View mission
implementationChecked Sep 25, 2026

RedPajama: Open Pretraining Datasets for Reproducible Foundation Models

Vipul Ved Prakash · Together AI

Co-led the open-source release of RedPajama, a 1.2-trillion and 30-trillion token pretraining dataset replicating the LLaMA pretraining corpus to enable transparent, fully reproducible foundation model research.

Scope & limitations

Web-scale data deduplication and filtering at 30T scale require petabyte-scale distributed compute and risk residual benchmark contamination.

Context: 30T token web corpus, deduplication pipelines, MinHash LSH, toxicity filtering.

View mission
explanationChecked Sep 25, 2026

Responsible AI by Design: Operationalizing Governance in the Fortune 500

Lan Guan · Accenture

Formulated the enterprise compliance and ethical AI operating manual, translating global regulatory standards (EU AI Act, NIST AI RMF) into actionable engineering validation gates and bias mitigation pipelines.

Scope & limitations

Subjective regulatory interpretations and jurisdiction-dependent compliance mandates create friction for multinational AI model rollouts.

Context: Algorithmic bias audits, EU AI Act compliance tiers, automated red-teaming validation gates.

View mission
Showing 1–16 of 567 proofs of work