Cohere

for the development and deployment of their products. [ 14 ] [ 15 ] In June 2022, Cohere launched Cohere Labs (previously called Cohere For AI), a nonprofit research lab and community to contributing open-source, fundamental machine learning research. It was led by Sara Hooker , a former research scientist at Google Brain [ 16 ] until her departure in September 2025. Cohere Labs is now led by Marz

nership rooted in customer success, now and for the future Through the process, Cohere provided hands-on benchmarking and experiments to find the right configuration, helping solidify Cohere as a trusted retrieval partner for Ada. Says Melidis, “Whether we were using SageMaker or Vault, we felt like there wasn't a strong pressure. The teams just wanted us to be successful.” The Ada team has also s

f answer a financial services company gives a banking customer will be different than the one an airline gives a traveler whose flight was just cancelled. But before AI agents can generate the right answers, they must find the right information. That’s where retrieval comes in. Every customer interaction begins with retrieval, giving AI agents the information they need to resolve conversations. “I

026 4 minute read Ada finds the right ingredients for faster, more efficient AI agents Key takeaways Accelerate AI agent retrieval: Achieved 77 ms median latency and 90.7% top-10 recall with Cohere Embed on Amazon SageMaker AI. Improve cost efficiency: Delivered 2.7x greater throughput-per-dollar for offline embedding with Cohere Model Vault. Reduce storage by 4x: Cut vector storage requirements w

VPCs Security Protect your data at every stage See deployment options By Industry Financial Services Public Sector Technology Telecommunications Energy and Utilities Healthcare Life Sciences Manufacturing Featured Model Vault provides fully-isolated, performant inference with Saas simplicity Insights Customer Stories For Developers Developers Models Overview Docs Discord LLM University Connect Par

red CoreWeave’s support staff and engineers to deliver faster, higher-quality support, paving the way for further agentic AI adoption. The promise of AI agents to help companies improve their customer service is clear. So far, though, only a few organizations have turned that vision into reality. CoreWeave , a global AI cloud leader based in the US, is acutely aware that solving customers’ problem

Ada's scale — thousands of customer knowledge bases, some containing tens of thousands of articles — results quickly compounded. The team reduced its vector storage footprint by 4x with no measurable retrieval degradation. "It was literally a free win,” says Melidis. Better retrieval gave Ada's agents better information to work with, supporting its north star metric: successful resolution. A part

alified to solve their issues. It has a simple motto: Customer first, always . “Every interaction has to protect their time and keep their systems running” John Mancuso, VP of Customer Experience, CoreWeave The initial challenge Facing rapid growth in demand, CoreWeave wanted to streamline its support system by reducing the number of third-party tools, increasing automation, and reducing the amoun

latency budget that left little room for retrieval Scale: Perform across regions at multi-tenant scale A structured approach to find a solution Ada conducted a structured evaluation to find a solution for both offline and online use cases. "We started in a very structured way—a four-way bakeoff,” explains Melidis. The bakeoff, a systematic, time-bound testing period, vetted each approach against b

ndustries Why leading teams trust Cohere technology Ada finds the right ingredients for faster, more efficient AI agents Read more technology How CoreWeave used Cohere North to transform its customer support in 90 days Read more technology Integrating Cohere into Draftwise boosted the quality of search results by 30% Read more The team at Cohere has made foundational contributions to generative AI

Tube CFO Martin Kon as president" . The Globe and Mail . Retrieved October 3, 2023 . 1 2 "Google Cloud teams up with NLP startup Cohere on multiyear partnership around TPUs" . TechCrunch . November 17, 2021 . Retrieved December 31, 2021 . 1 2 Cohere (September 7, 2021). "Cohere Raises $40 Million in Series A Financing to Make Natural Language Processing Safe and Accessible to Any Business" . Globe

Small Translate Machine translation model Parse New Document parsing model Embed New Semantic representation model Rerank New Retrieval optimization model Models Overview Product Products Overview Total Cost of AI Ownership Pricing Featured North 2: Cohere's latest platform update — with 15+ new features Deploy Model Vault Dedicated model inference platform Private Deployments On-prem or isolated

flexibility, performance, and operational ease. Cohere Embed could fit into its existing architecture, allowing the team to optimize retrieval while preserving the models and infrastructure already supporting its agents. The project gave Ada more than a point solution. It established a retrieval architecture that the team can continue to evolve as its AI stack changes. With Cohere already prepari

f retrieval is slow, our agent is slow and everything that stems from it... and if retrieval is wrong, then the agent is wrong as well,” explains Christos Melidis, PhD, Staff Machine Learning Scientist at Ada. The initial challenge Ada’s online use case: every millisecond counts Now imagine you are calling in, asking an AI agent a question and hearing an unnatural pause before every answer. A few

as geographically closer to Ada’s services, it was the sensible choice for the latency-sensitive live conversations. With Cohere Embed on SageMaker AI, Ada achieved median retrieval latency of 77 milliseconds (ms) — only 0.077 seconds for that step in the response process. For a caller, shaving time from retrieval helps reduce the silence between asking a question and hearing the agent respond. Fo

etter it is.” In other words, when voice enters the chat (literally), the stakes are higher. Three requirements Within the online use case, Ada set a tighter latency budget for voice than its previous retrieval setup could deliver consistently. The solution had to meet three requirements: Quality: Find the right information in each customer's knowledge base Latency: Fit within an end-to-end voice

cal representations that make it possible to search for semantically similar information. “We saw better performance, better quality in the answers, and a better experience for the clients.” Christos Melidis, PhD, Staff Machine Learning Scientist, Ada The recipe for success emerges In the online use case, Cohere Embed quickly emerged as a star, achieving 90.7% top-10 recall. Because SageMaker AI w

mocratize natural-language AI" . Fast Company . Archived from the original on December 22, 2021 . Retrieved December 31, 2021 . 1 2 3 4 Goldberg, Natalie (June 8, 2023). "Nvidia invests in Google-linked generative A.I. startup Cohere" . CNBC . Archived from the original on October 6, 2023 . Retrieved October 6, 2023 . 1 2 Silcoff, Sean (September 7, 2021). "Toronto AI startup Cohere raises US$40-m

th whom Gomez had done research at FOR.ai. [ 10 ] [ 11 ] All of the co-founders attended the University of Toronto . [ 12 ] Gomez is the company's CEO. [ 11 ] In December 2022, Martin Kon, the former CFO of YouTube , became president and COO. [ 13 ] In November 2021, Google Cloud announced they would help power Cohere's platform using their infrastructure, and Cloud's TPUs would be used by Cohere

e(s) of Work How will AI change the way we work? Aya Models Multilingual AI at scale All Papers Initiatives Research Scholars Finding the new generation of ML talent Open Science Community Championing global, open science Catalyst Grants Supporting impactful ML endeavors Resources Blog Hugging Face Events Featured The future of work debate has an evidence problem About Careers Newsroom Customer st

gn on to a voluntary White House commitment on measures for testing, reporting, and research on the risks of AI. [ 18 ] On September 27, 2023, it was announced that Cohere had also signed Canada's voluntary code of conduct for AI, to promote the responsible development and management of advanced generative AI systems. [ 19 ] In 2024, Cohere opened an office in New York City. [ 20 ] [ 21 ] [ 22 ] I

s fast is fundamental to success. The pioneers who rely on CoreWeave’s infrastructure and support include the world’s top AI labs and global enterprises. Their time is highly valuable; just an hour of downtime can be extremely expensive. That’s why CoreWeave has built its support culture on the principle of "Direct to Expert," a commitment to connecting customers quickly with the engineers most qu

Enterprise AI: Private, Secure, Customizable | Cohere Skip to content Rerank 5 is here: sharper search, fewer tokens. Now with Context Focus. Learn more Products Solutions Resources Blog Research Company Sign in Request a demo Platform North Enterprise-ready AI for business Compass Intelligent search and discovery Models Command Generative language models Transcribe Speech recognition model North

r offline workloads, Model Vault delivered approximately 2.7x greater throughput-per-dollar than Ada's self-hosted setup. Its managed autoscaler handled republishing spikes faster than Ada's self-managed scaling. Across the retrieval system, Embed enables compression without performance impact. Because Embed supports compression natively, Ada did not need to build its own quantization pipeline. At

025 . ↑ Satariano, Adan (April 24, 2026). "A.I. Start-Ups From Canada and Germany Merge to Take On Silicon Valley" . The New York Times . Retrieved April 24, 2026 . ↑ Scott, Josh (September 8, 2021). "Cohere raises $40 million USD to make natural language processing more accessible" . BetaKit . Retrieved December 31, 2021 . ↑ Sullivan, Mark (September 7, 2021). "Ex-Googlers raise $40 million to de

ieh Fadaee. [ 17 ] In December 2022, Cohere released a multilingual model for understanding text that would work with over 100 languages, to help users search for documents by meaning instead of keywords. This type of process was not previously widely available in languages other than English. [ 13 ] On September 12, 2023, it was announced that Cohere had become one of fifteen tech companies to si

ur rules. North by Cohere is a complete agentic enterprise AI platform that you control and deploy on your terms. Trusted by industry leaders and developers worldwide AI for Empowerment Your freedom. Your focus. Cohere keeps it that way. Learn how AI helps you reclaim time for what truly moves you. Explore now Purpose-built AI solutions north platform North: Enterprise AI Without Compromises A ful

re AI has flown under the radar. That may be about to change" . Venture Beat . Archived from the original on October 6, 2023 . Retrieved October 3, 2023 . ↑ "AI startup Cohere raises US$40-million for natural language software: Globe and Mail" . University of Toronto News . September 8, 2021 . Retrieved December 31, 2021 . 1 2 3 O'Kane, Josh (December 14, 2022). "Rising AI startup Cohere hires You

ting reveals a key insight: two use cases, two solutions For offline knowledge-base indexing, Ada uses Cohere Model Vault . Dedicated, Cohere-managed infrastructure runs models privately and at scale, and supports bulk embedding and ingestion. For its online retrieval path, Ada self-hosts Cohere Embed on dedicated Amazon SageMaker AI instances. Cohere Embed converts content and queries into numeri

milliseconds might sound insignificant on paper, but inside a conversation, you feel it. A fast retrieval model gives customers more confidence that their questions will be answered in a timely manner instead of wondering whether the call dropped. “If you're on the phone, you don't have a visual. You don't see an indicator blinking. You're hearing silence,” explains “The smaller the silence, the b

AI agent retrieval: How Ada scales with Cohere | Cohere Skip to content Rerank 5 is here: sharper search, fewer tokens. Now with Context Focus. Learn more Products Solutions Resources Blog Research Company Sign in Request a demo Platform North Enterprise-ready AI for business Compass Intelligent search and discovery Models Command Generative language models Transcribe Speech recognition model Nort

oth of Ada's workloads. The bakeoff’s ingredients: 15,000 evaluation queries Online and offline workloads Four embedding approaches Three requirements: quality, latency, and scale One north star: successful resolution From the start, the testing challenged one of Ada’s assumptions: that a single deployment could efficiently serve both workloads. Two use cases demanded two technology solutions. Tes

5 minute read How CoreWeave used Cohere North to transform its customer support in 90 days Key takeaways CoreWeave transformed its customer support within 90 days by embedding Cohere’s North directly into its Slack-based workflows . North’s automated triage and context-gathering improved routing accuracy, accelerated resolution times, and supported strong customer satisfaction scores. North empowe

tners Events Webinars Merch Store Featured Ada finds the right ingredients for faster, more efficient AI agents Blog The latest news, launches, and insights Read more Cohere to acquire AISLE to advance cyber-resilient sovereign AI North 2: Enterprise AI without compromises Introducing Embed 5—A new family of frontier embedding models Research Cohere Labs Cohere’s ML research lab Explorations Futur

ly ownable agentic platform built to bring peak security, intelligence, cost efficiency, and control together. Explore North cohere models Our models. Your business. Tools for generative AI, advanced search, and multilingual capabilities—built to keep your enterprise in full control. Explore models model vault Scalable, secure AI deployment Model Vault provides fully-isolated performance inference

ories Enterprise AI case studies and success stories See how our customers and partners streamline operations, deliver intelligent user experiences, and transform business outcomes with enterprise-ready AI. Request a demo Ada finds the right ingredients for faster, more efficient AI agents How CoreWeave used Cohere North to transform its customer support in 90 days Fujitsu launches Takane AI model

in partnership with Cohere Draftwise boosts legal productivity with Cohere's AI models BambooHR optimizes AI search and retrieval Oracle launches over 100 generative AI use cases with Cohere Atomicwork builds Atom AI with Cohere