Adept: a machine learning research lab Startup: AdeptRecommended by: Niki Pezeshki, FelicisRelationship: No financial interestTotal funding: $415 million, according to the companyWhat it does: Adept is a machine-learning research and product lab aiming to build AI technology that can automate human tasks as varied as running software and browsing the web.Why it's on the list: "Fresh off a massive round of funding, Adept is poised to deliver artificial general intelligence that truly benefits working people," Pezeshki said. "So much software is repetitive tasks, and Adept is creating an AI that will enable workers to increase productivity and focus on more creative tasks." AlphaSense: a business-data-intelligence platform Startup: AlphaSenseRecommended by: James Luo, CapitalGRelationship: InvestorTotal funding: $623 million, according to PitchBookWhat it does: AlphaSense aggregates over 10,000 qualitative and quantitative data sources in one platform, with an intuitive search interface and generative-AI capabilities that allow users to quickly find and summarize insights across content, including earnings calls, financial reporting, expert-call transcripts, industry journals, and research vendors.Why it's on the list: "AlphaSense has built out significant infrastructure for utilizing generative AI and is one of the first companies to realize these capabilities in a commercial capacity," Luo said. In April, the company announced it had raised $100 million from investors, including Viking Global and affiliates of Alphabet and Goldman Sachs. And more recently, the company was in talks to raise $150 million more in funding this year at a $2.5 billion valuation, Bloomberg reported. Captions: an intelligent video editor Startup: CaptionsRecommended by: Ev Randle, Kleiner PerkinsRelationship: InvestorTotal funding: $40 million, according to the companyWhat it does: Captions is an intelligent video editor that allows creators to edit, caption, dub, and add music or effects to their videos with a few taps.Why it's on the list: Caption's cap table features the crème de la crème of venture funds, including Kleiner Perkins, Sequoia Capital, Andreessen Horowitz, and SV Angel. Founded by Gaurav Misra, the former head of design engineering at Snap, Captions has hooked investors with its domain expertise and an exploding daily active user base."Captions has emerged as the leading AI-powered platform for video content," Randle said. "Millions of users have already leveraged it to tell their stories and engage audiences." CentML: optimizes machine learning models Startup: CentMLRecommended by: Rares Crisan, GGV CapitalRelationship: No financial interestTotal funding: $3.5 million, according to the companyWhat it does: CentML's technology optimizes machine-learning models so they work smarter and faster, which then helps cut computing costs. Why it's on the list: The Toronto-based startup came out of stealth in 2022 with a seed round led by the AI-focused firm Radical Ventures, positioning itself as a viable alternative in light of the global shortage of GPUs — the chips needed to train and run AI models.Its cofounder and CEO, Gennady Pekhimenko, leads the Efficient Computing Systems lab and is an assistant professor at the University of Toronto, and is also a faculty member at the Vector Institute, a Canadian not-for-profit AI-research institute. Crisan says CentML reduces the cost of training and inference by 60 to 90%. "We believe CentML will benefit from the proliferation of AI models that need training and maintenance over the coming years," he added. Character.AI: AI-powered chat Startup: Character.AIRecommended by: Ev Randle, Kleiner PerkinsRelationship: No financial interestTotal funding: $193 million, according to the company.What it does: Character.AI allows users to chat with a multitude of AI-powered characters, ranging from Taylor Swift to William Shakespeare. In March, Character.AI landed a $150 million Series A from Andreessen Horowitz, nabbing a $1 billion valuation.The company, which was also on Insider's most promising generative-AI startup list, has created a strong competitive moat around itself, setting itself apart from competitors by building on top of third-party models, some investors say. Why it's on the list: 'It's an incredibly engaging technology," said Randle. Durable: AI for business services Startup: DurableRecommended by: Josh Coyne, Kleiner PerkinsRelationship: No financial interestTotal funding: $6.25 million, according to PitchBookWhat it does: Durable uses AI to quickly build websites and other business services.Why it's on the list: Running a business can be difficult, and Durable is gaining traction by using AI to help companies quickly build websites and manage other business functions in a matter of seconds."In practice, this takes the form of a suite of different tools (e.g. website builder, CRM, marketing automation, invoicing, spend management, AI copilot) purpose-built to help business owners focus on what they do best — generating revenue — and offloading the rest," Coyne said.The startup raised a $6.25 million seed round at the end of 2022. Entos: using AI to design and discover drugs Startup: EntosRecommended by: Tiba Aynechi, Norwest Venture PartnersRelationship: No financial interestTotal funding: $56 million, according to PitchBook.What it does: Entos uses generative AI and machine learning to design and discover drugs. Why it's on the list: "Entos has used their NeuralPlexer platform to advance multiple drug candidates from hit identification to preclinical studies in under two years, a rapid pace for what is typically six or more years," Aynechi said. "They are planning to enter clinical studies soon with a differentiated lead candidate that has been optimized by their AI-driven multi-parameter platform." Foundry: a computing and infrastructure platform for AI workloads Startup: FoundryRecommended by: Jill Chase, CapitalGRelationship: No financial interestTotal funding: UndisclosedWhat it does: Foundry is a simplified computing and infrastructure platform for AI workloads.Why it's on the list: "With the ever rising interest in AI today, the hardware infrastructure that underpins AI use cases and workloads has proven hard to come by and remarkably difficult to use properly," Chase said. "Foundry is attacking this problem by creating a simplified, low friction compute and infrastructure platform that developers can use to instantly and optimally deploy their AI workloads." GPTZero: detects AI-generated text Startup: GPTZeroRecommended by: Asheem Chandna, GreylockRelationship: No financial interestTotal funding: $3.5 million, according to the companyWhat it does: GPTZero analyzes text and determines if it was AI-generated or written by a human. Why it's on the list: Initially billing itself as an anti-cheating detector for educators, GPTZero has since expanded its vision for a much broader range of possibilities for the need to detect AI-generated text. "Detecting AI-generated voice, video, text, etc. will be critical to preventing fraud," Chandna said. "LLMs have been the story of 2023, and while they offer many benefits, understanding when content has been produced by them is useful to many different businesses." Hugging Face: an open-source and community-driven AI platform Startup: Hugging FaceRecommended by: Dan Cahana, GGV CapitalRelationship: No financial interestTotal funding: $164.9 million, according to the company What it does: Hugging Face is an open-source and community-driven AI platform, which provides tools to help users build machine-learning models and datasets.Why it's on the list: Hugging Face has quickly become one of the biggest open-source communities for those seeking to build large language models and has been named "the GitHub of AI." In addition to hosting these AI models, it offers the tools to make them better. "The way the team has fostered the open-source community is very exciting.With Hugging Face, users have access to more than half a million shared AI models in an effort to democratize AI and AI building," said Cahana.Founded in 2016, that startup has raised millions from VCs including Lux Capital, Sequoia, and Lee Fixel's Addition, reaching a $2 billion valuation in the process. It was also in talks to raise at least $200 million at double that value, Forbes reported. LangChain: a framework for building applications using LLMs Startup: LangChainRecommended by: Navin Chaddha, Mayfield FundRelationship: No financial interestTotal funding: $10 million, according to PitchBookWhat it does: LangChain has created a framework for developing more complex applications powered by LLMs.Why it's on the list: LangChain launched last October as an open-source project by Harrison Chase while he worked as an engineer at the machine-learning startup Robust Intelligence. The startup scored a $10 million seed round earlier this year — led by Benchmark — and it raised another round between $20 million and $25 million, led by Sequoia. "The company is very early, but I am nominating it based on the problem it is solving and the quality of the founding team," Chaddha said. Leena AI: an AI-powered virtual work assistant Startup: Leena AIRecommended by: Anant Vidur Puri, Bessemer Venture PartnersRelationship: InvestorTotal funding: $40 million, according to PitchbookWhat it does: Leena AI is an AI-powered virtual work assistant designed for companies.Why it's on the list: "Equipped with Leena AI's proprietary large language model, WorkLMTM, Leena AI empowers enterprises worldwide to redefine how employees engage with work, delivering a transformative impact on productivity and efficiency," said Puri. "Leena AI empowers employees with a versatile toolset to accomplish tasks with exceptional precision and speed." LlamaIndex: LLM-powered search for large databases Startup: LlamaIndexRecommended by: Jerry Chen, Greylock PartnersRelationship: InvestorTotal funding: $8.5 million, according to PitchBookWhat it does: LlamaIndex provides the tools for developers to use LLMs for searching large, private, or proprietary databases. Why it's on the list: With LlamaIndex, businesses can use LLMs to search and generate results from their own internal proprietary data, such as emails, documents, databases, and more. Greylock Partners recently led LlamaIndex's $8.5 million seed round. "It is the key technology to make LLMs useful for enterprise apps," Chen said. Luma AI: an app to create 3D imagery Startup: Luma AIRecommended by: Lucy Deland, Inspired CapitalRelationship: No financial interestTotal funding: $25.5 million, according to the companyWhat it does: Luma AI lets users create realistic 3D images on their phones using AI.Why it's on the list: Thanks to the AI boom, more companies than ever are leveraging the technology to expand the average user's ability to create visual effects. Luma, which raised a $20 million Series A this spring, is no different."Luma's vision is an incredibly compelling use case for AI," Deland said, adding that the startup was "adding a new dimension to how we experience photos and video.""Having cofounded Paperless Post," she said, "I've long been interested in the interplay between digital and real-world mediums and think Luma's approach is an exciting one." Lumachain: a real-time platform for food supply chains Startup: LumachainRecommended by: Tess Hatch, BessemerRelationship: InvestorTotal funding: $22 million, according to PitchBookWhat it does: Lumachain's real-time platform for food supply chains tracks the origin, location, and condition of goods. Why it's on the list: Food production is one of the oldest industries in the world, but it hasn't been touched by software — until Lumachain came in, Hatch said. The startup raised $19.5 million in Series A funding last year and has netted some big-name clients in the food-processing industry. "The company is already working with the biggest meat processing companies, including JBS and Tyson, and is on track to 5.5 times ARR this year," Hatch said. Magic: an AI teammate for software engineering Startup: MagicRecommended by: Jill Chase, Capital GRelationship: InvestorTotal funding: $193 million, according to the companyWhat it does: Magic is developing an AI teammate for software engineering that allows developers to communicate in natural language with the AI to complete code faster. Far beyond simple code completion or next-line prediction, Magic is building an AI colleague that acts as a pair to a programmer and is able to understand and continually learn more about the context of the work and user.Why it's on the list: "Magic's incredible research on neural memory and reinforcement learning allows for a truly differentiated product that will unlock significant value and open up a new market of use cases," said Chase. "Specifically, in June of this year, Magic announced their proprietary LLM for code LTM-1, which is a nLLM with a 5 million token window (an order of magnitude larger than the next closest LLM context window). LTM-1's massive context window allows for higher quality generations than other LLMs, which is particularly valuable for writing code, where LTM-1 can see a user's entire code repository when generating suggestions." Mezli Startup: MezliRecommended by: Jo Zhu, Pear VCRelationship: No financial interest Total funding: $3.63 million, according to PitchbookWhat it does: Mezli is a fully autonomous restaurant that serves Mediterranean-inspired grain bowls starting at $6.99 in San Francisco's food-truck park Spark Social. The restaurant is a blue-and-white, windowless shipping container that's filled with robots putting together meals with pre-made ingredients that customers order via a touchscreen menu.Mezli's three cofounders, Alex Kolchinski, Alex Gruebele, and Max Perham, were Stanford graduate student engineers who were inspired by the lack of quick, affordable, and healthy food options nearby. Why it's on the list: Robots appearing on the food scene is not entirely new, Zhu said. "But usually they are with low-fidelity and low-movement robots for coffee or smoothies (not requiring grabbing, chopping, and finer-tuned muscle finesse like picking up tomatoes)," she added. What sets Mezli apart is that it operates entirely with robots, with little human help to prepare the food. Mindee: document processing automation Startup: MindeeRecommended by: Tiffany Luck, GGV CapitalRelationship: InvestorTotal funding: $23.4 million, according to PitchBookWhat it does: Mindee helps developers avoid manual data entry with an API that lets them turn raw data in a paper document into usable, structured data.Why it's on the list: Founded in France, Mindee has been on a tear since exiting stealth mode in October of 2021. It expanded to the North American market and opened new offices in San Francisco to support its growing staff. Luck, who led Mindee's $14 million Series A round, said the company is catching a wave of workflow automation. "Despite the rapid migration toward AI and automation, key information from invoices, expense receipts, compliance documentation, etc. still requires manual entry into data systems," Luck said. "Mindee eliminates the need for manual data entry and provides real-time, human-level accuracy in data extraction from paper and digital documents." Next Insurance: insurance for small businesses Startup: Next InsuranceRecommended by: Jesse Wedler, CapitalG Relationship: InvestorTotal funding: $881 million, according to the companyWhat it does: Next Insurance's platform provides small businesses with insurance coverage. Why it's on the list: "Small businesses have been underserved historically by traditional insurance carriers who are focused on landing contracts with large customers, leaving this large tail market without access to the right coverage," Wedler said. "Next Insurance leverages automation and AI to make the insurtech business model work for this small business segment." Orby AI: AI for repetitive work tasks Startup: Orby AIRecommended by: Arash Afrakhteh, Pear VCRelationship: InvestorTotal funding: $4.5 million, according to the company. What it does: Orby's generative-AI technology observes a user's activities, identifies repetitive work steps, and automatically generates code that automates those tasks.Why it's on the list: Orby has a "world-class team that has been executing on their vision of automating any repetitive tasks, leaving the innovation to humans and performing the mundane by machines," Afrakhteh said. "Large enterprises are adopting their robust AI solution." Pinecone: an AI vector database Startup: PineconeRecommended by: Sai Senthilkumar, Redpoint Ventures Relationship: No financial interestTotal funding: $138 million, according to PitchBookWhat it does: Pinecone is an AI vector database for capturing and storing representations of unstructured data such as audio, images, and video.Why it's on the list: Senthilkumar named Pinecone "the database for the AI revolution." As AI and machine-learning models rapidly become more powerful, developers are feeding vastly more complex datasets into their models. Storing and managing these massive data assets, which don't necessarily fit easily into traditional databases, has become a significant chokepoint for many startups. Pinecone's products are designed to help developers deploy complex machine-learning applications at scale with ease. "It serves a crucial component of the next-generation AI infrastructure stack, fueling a range of use cases from object recognition to recommendation systems," Senthilkumar said. Poly: a generative-AI startup for design assets and creative textures Startup: PolyRecommended by: Ryan Isono, Felicis VenturesRelationship: InvestorTotal funding: $3.9 million, according to PitchBookWhat it does: Poly is a web-first generative-AI startup for design assets and creative textures. Why it's on the list: "Poly has seen exponential growth this year and is being used by designers across the spectrum from interior design to gaming," Isono said. "They're working on more exciting features to become the leading platform for creating various design assets from a simple prompt." Predibase: a platform for quickly building and deploying custom machine-learning models Startup: Predibase Recommended by: Niki Pezeshki, Felicis VenturesRelationship: InvestorTotal funding: $28.4 million, according to the company What it does: Predibase is a low-code AI platform for quickly building and deploying custom machine-learning models and LLMs. Why it's on the list: Predibase came out of stealth last year with a $16.25 million funding round led by Greylock and in May extended that round with an additional $12.2 million, which Felicis led. The startup's founding team comes from big tech companies such as Uber, where the cofounders Piero Molina and Travis Addair were heavily involved in developing the open-source technologies that Predibase was built on top of. Its other cofounders include Devvret Rishi, who was a Google Cloud AI product manager, and Chris Ré, a Stanford computer science professor who cofounded Lattice.io, a machine-learning company Apple bought in 2017. Pezeshki says that Predibase has had a great year. "They launched new features this year, including a data-science copilot that helps devs improve their models, private hosting of custom LLMs, and a two-week free trial," he said. Replicant: AI customer service Startup: ReplicantRecommended by: Chester Ng, AtomicRelationship: InvestorTotal funding: $113 million, according to PitchBook. What it does: Replicant uses AI to replicate a human customer-service agent and engage in natural and nuanced conversations to resolve customer-support issues. Why it's on the list: "During this seminal moment for AI, getting customers off being on hold forever is definitely what AI should be used for — saving costs for businesses while increasing their customer satisfaction," Ng said. "Replicant is transforming the $1.3 trillion customer service industry by solving an urgent business need with proven ROI, brand-name customers, and an impressive growth trajectory." Replicate: machine-learning models on the cloud Startup: ReplicateRecommended by: Aydin Senkut, Felicis VenturesRelationship: No financial interestTotal funding: $17.9 million, according to PitchBookWhat it does: Replicate runs machine-learning models on the cloud with just a few lines of code.Why it's on the list: With Replicate, a company doesn't need a team of experienced engineers to harness the power of AI and machine learning — and that promise has helped the startup sign up a "huge number of users in mere months," Senkut says."Replicate lets you run machine-learning models with a cloud API without having to understand the intricacies of machine learning or manage your own infrastructure," he said. "You can run open-source models that other people have published or package and publish your own models." Run:ai: software to boost GPU performance Startup: Run:aiRecommended by: Lonne Jaffe, Insight PartnersRelationship: InvestorTotal funding: $118 million, according to PitchBookWhat it does: The Tel Aviv-based Run:ai helps companies do more with their graphics-processing units, allowing engineers to perform complex tasks, such as training LLMs more quickly and for less. Why it's on the list: As the AI revolution continues, many companies are scrambling to amass the computing power needed to train and run AI models. And Run:ai's software helps companies beef up that computing power. The company recently announced a partnership with Nvidia, the world's leading producer of GPUs. "With the increasing use of GPUs for AI systems, the global GPU shortage and rationing, inflationary pressures, and the need for companies to reduce spend, there is extraordinary pent-up demand for their product," Jaffe said. SaaS Labs: an automation platform for sales and support agents Startup: SaaS LabsRecommended by: Anant Vidur Puri, Bessemer Venture PartnersRelationship: No financial interestTotal funding: $99.5 million, according to PitchBookWhat it does: SaaS Labs is an automation platform for sales and support agents at small and medium-sized businesses.Why it's on the list: With its flagship products, JustaCall, Helpwise, and Callpage, SaaS Labs is now catering to over 7,000 customers across the world. Its primary focus of catering to the call and contact center software market is in a massive market that's "growing at a healthy rate," says Puri.The startup is cash efficient and growing steadily, and is in a unique spot to benefit from the ongoing AI adoption among companies, he added. "They have all the reasons to win in 2023 — thousands of existing customers, clear understanding of use cases, lots of useful data to fine-tune models for different industries, ample amount of capital, and a great team." Secureframe: automated security and compliance services Startup: SecureframeRecommended by: Josh Coyne, Kleiner PerkinsRelationship: InvestorTotal funding: $79 million, according to the companyWhat it does: Secureframe offers automated security and compliance services with AI.Why it's on the list: Secureframe is well-positioned to tackle the "deceptively large market" of cybersecurity thanks to Shrav Mehta, the firm's seasoned founder who's worked on several unicorns such as Scale AI and Ramp."Secureframe has built an AI-powered platform that streamlines enterprise compliance adherence while continuously monitoring security postures and automatically remediating any gaps or vulnerabilities," Coyne said. The startup raised a $56 million Series B last year. Treat: generative AI for product photos Startup: TreatRecommended by: Mike Duboe, GreylockRelationship: InvestorTotal funding: $8.5 million, according to the companyWhat it does: Treat uses generative AI to make customized product images for e-commerce brands based on targeted demographic data to create ads. Why it's on the list: "In the face of sweeping changes and headwinds across ad platforms over the past two years, great creative and accurate measurement are the last remaining sources of alpha for advertisers," Duboe said of Treat's specific combination of customer analytics and generative AI.Treat has already claimed some noteworthy direct-to-consumer brands as customers, including the trendy cookware startup Caraway and the salon-quality hair-care brand Briogeo. Earlier this year, the startup raised an $8.5 million seed round, which Greylock Partners led. Twelve Labs: helps developers build video applications Startup: Twelve LabsRecommended by: Tiffany Luck, GGV CapitalRelationship: No financial interestTotal funding: $17 million, according to the companyWhat it does: Twelve Labs helps developers build video applications that can search for specific people, objects, or scenes in a video, using everyday language.Why it's on the list: The company's tech isn't too dissimilar from the Google tech that recommends videos across Google Search and YouTube. But while Google has kept this tech to itself, Twelve Labs is among the first to bring it to market. Two years into the business, it has paying customers, a growing headcount, and $12 million in fresh funding after the startup closed a seed extension round in December of 2022."In the era of AI, content creation is exploding," Luck said. "Twelve Labs, with their suite of APIs, provides next-generation video search infrastructure, making video content searchable and understandable." WellSaid: AI for corporate presentations and ad campaigns Startup: WellSaidRecommended by: Edward Yip, NorwestRelationship: No financial interest Total funding: $12 million, according to the companyWhat it does: WellSaid uses AI to create corporate training presentations, advertising campaigns, and more.Why it's on the list: When accounting for recording, voice acting, and post-production, both internal and external company presentations can be costly and time-consuming. But WellSaid is using AI to cut down on the time and money traditionally associated with training presentations, advertising campaigns, and other use cases. Yip said the startup is a "top-performing product" in the space. Read the full article here