AUTHOR
Ravi Kapoor
Models & Infrastructure Editor
Ravi follows foundation models, inference platforms and the rise of regional AI ecosystems. His work examines performance alongside language coverage, deployment control and cost.
Email Ravi →Experience and editorial focus
He has covered cloud infrastructure and open-source software across multiple technology cycles.
Areas of expertise
- Foundation models
- Open-source AI
- Inference platforms
LATEST WORK
Articles by Ravi Kapoor
NVIDIA and Palantir bring sovereign AI to critical supply chains
The companies are combining custom Nemotron open models, Palantir Foundry and AIP, and a cloud or on-premises reference architecture, beginning with NVIDIA's own supply chain.
9 min read →AI ModelsOpenAI launches GPT-6 Astra for ChatGPT Work, Codex and the API
OpenAI's newest flagship model targets computer use and professional workflows, with API pricing beginning at $10 per million input tokens and $50 per million output tokens.
10 min read →IndustryNVIDIA partners target up to 2 gigawatts of Australian AI infrastructure by 2027
Eight infrastructure partners plan capacity for NVIDIA DSX systems, pairing sovereign compute ambitions with new demands for power, cooling and network capacity.
9 min read →IndustryMistral raises €3 billion as Samsung backs Europe’s open-weight AI push
The French AI company says the round values it above €21 billion and will fund model research, compute, infrastructure and international expansion.
9 min read →IndustryOpenAI and Thailand launch an eight-week AI startup accelerator
Ten Thai health, wellness and education startups will receive mentoring, evaluation support and $2,000 in API credits through a new public-private accelerator.
9 min read →IndustryAWS and NVIDIA plan two million additional GPUs for AI infrastructure
The companies say the 2027-2028 expansion will span Blackwell Ultra, Rubin and Rubin Ultra systems, Vera CPUs, networking, open models and robotics.
9 min read →IndustryOpenAI launches commercial operations in Brazil
A São Paulo team will work with businesses, developers, researchers and public institutions as OpenAI expands partnerships and training programs in Brazil.
8 min read →IndustryGoogle selects 28 AI startups focused on energy systems
The North American and European cohorts span grid software, geothermal drilling, autonomous maintenance, data-center power access and demand flexibility.
8 min read →IndustryOpenAI publishes first results for its Jalapeño inference chip
OpenAI says its first custom inference accelerator delivers lower latency and more work per watt across three open-weight models, marking a deeper move into AI infrastructure.
8 min read →IndustryNVIDIA pitches NVLink Fusion as the bridge for custom AI chips
NVIDIA says hyperscalers can combine custom XPUs with its scale-up networking, rack architecture and operations stack instead of rebuilding every layer around a new accelerator.
8 min read →Model PlatformsNVIDIA puts Groq 3 LPX into production for agent inference
The Groq 3 LPX extension to Vera Rubin targets the token-generation stage of long-context agents, while NVIDIA reports early adoption by Nebius, CoreWeave and SpaceXAI.
8 min read →AI ModelsGPT-5.6 arrives across ChatGPT, Codex and the OpenAI API
OpenAI's July flagship release brings one model family to consumer chat, coding agents and the API, followed by substantial price cuts for its Luna and Terra variants.
7 min read →Research & DataMachine learning algorithms: how to choose the right family for a real problem
Regression, trees, clustering and neural networks solve different kinds of problems. This practical map starts with the decision, the data and the cost of being wrong—not a list of fashionable algorithms.
13 min read →Research & DataTypes of machine learning models: a practical guide from regression to transformers
Model types are easier to understand when organized by what they learn, what data they consume and how their output will be used. Here is a practical taxonomy for choosing and evaluating them.
12 min read →Model PlatformsThe model race is becoming local, multilingual and sovereign
Sarvam’s open 30B and 105B models ignited a broader conversation about who builds a country’s AI layer—and whether global benchmarks capture the languages and workflows that matter locally.
7 min read →AI ModelsWhat an open-model breakthrough actually changes
A strong benchmark score can earn attention overnight. The harder questions are whether a model can be inspected, deployed affordably and trusted on the tasks that matter outside a leaderboard.
8 min read →IndustryAI hardware gets a second chance at the edge
The next AI device wave is less interested in novelty gadgets and more focused on on-device models, robotics and specialized computers that can respond without a permanent cloud round trip.
8 min read →AI ToolsSeven open-source AI projects worth knowing
The most useful open AI stack is larger than any one model. These seven projects cover model access, local runtime, high-throughput serving, routing, interfaces and agent workflows.
9 min read →AI ModelsSmall models are moving AI from the cloud onto devices
On-device language models trade frontier-scale breadth for low latency, privacy and zero per-request cloud cost. That makes them a different product primitive, not merely a smaller benchmark entry.
9 min read →