Led · Seed · Sep 2022

FindMine builds AI-driven styling and outfitting technology that generates outfits, shoppable content and editorialized product recommendations for retailers and brands. Its tech combines discriminative machine-learning models, vector databases and generative AI visualizations to understand products and create shoppable experiences from messy, unstructured brand data. The company says its technology directly powers $89 million in retailer and brand revenue annually and drives nearly $1 billion in downstream annual revenue through FindMine-generated content. Clients include Adidas (an early adopter), Lululemon, Lands’ End, DXL, Annie Bing, Vineyard Vines and French retailer CFOC, spanning the U.S., Canada, Europe and Japan. Founded in 2014 and taking on clients from 2016, FindMine emphasizes building its own models and infrastructure rather than relying on human-in-the-loop systems. Investors and partners cite the product’s ability to ingest unstructured data and scale outfit curation as key benefits for driving engagement, margin and inventory outcomes. FindMine builds predictive-intelligence and analytics technology that uses merchant sales data and AI to help brands predict trends, manage inventory, and automate content creation. The company’s platform generates highly-produced, curated editorial and visual shoppable assets in milliseconds for use across websites, campaigns, social media, emails and in stores. FindMine serves beauty, apparel and home brands, including Adidas, Perry Ellis, Reebok and Cole Haan. Leadership says the product improves top-line revenue via larger shopping carts and longer customer lifecycles while helping merchants improve profitability and gross margins through better inventory and logistics management. The firm positions its solution as a scalable alternative to one-to-one personalization by automating bespoke content at scale. FindMine plans to continue developing its AI-powered content engine and machine-learning capabilities and to use new data sets to better locate inventory along the supply chain and reduce split shipments. Its technology is marketed to help brands maintain visibility with customers and drive revenue and loyalty.

Participated · Series C · Apr 2022

Founded in 2015 and based in San Francisco, Lilt offers an AI-powered, human-in-the-loop translation platform that combines an AI translation engine with translator tools (hotkeys, style guides, custom term bases) to translate marketing, support and e-commerce content. The AI engine is regularly trained on fresh data and translator feedback to make recommendations while human translators retain final control. The platform supports around 40 languages, provides APIs and connectors (Slack, WordPress, GitHub, Salesforce, Zendesk, Adobe Marketing Cloud), and tracks job progress and translator hours. Lilt counts customers including Intel, Emerson, Juniper Networks and Orca Security across education, crypto, technology, defense and intelligence sectors, and has a workforce of over 150 people. The company plans to use the new capital to expand R&D, its engineering teams and customer footprint, and to add sales teams to match production teams in each region. Financially, Lilt recently raised a $55M Series C, bringing total capital raised to $92.5M. Lilt provides an AI-powered solution that combines adaptive neural machine translation technology, an enterprise translation management system, and professional translators. Its platform is designed to help organizations scale translation programs, accelerate time-to-market, and improve the global customer experience. Customers include Fortune 500 companies like Intel, consumer brands such as ASICS, and technology companies including WalkMe, Sprinklr, DigitalOcean, and Canva. Led by CEO and co-founder Spence Green, the company has opened additional offices in Indianapolis and Dublin, Ireland. Lilt has raised $37.5M in total funding and recently closed a $25M Series B. The company intends to use the capital to further develop its enterprise translation software, research natural language processing technology, and accelerate its go-to-market strategy. Lilt provides an interactive, adaptive machine translation system that continuously learns in real time from human feedback. Its approach combines human skills and machine learning to increase translation speed by up to five times while maintaining quality and accuracy. The technology is designed to help businesses localize materials into foreign languages and accomplish strategic objectives. Lilt positioned the offering for enterprise customers and government use through a newly announced partnership. CEO Spence Green said the partnership will help the company better serve customers with higher-quality translations more quickly. Lilt develops an interactive, machine-assisted translation platform that learns from human input to boost translator productivity. The platform evolved from a 2014 research prototype called Predictive Translation Memory, and includes predictive typing, interactive adaptive machine translation, automatic tag placement, a built-in termbase and an API. Co-founded by John DeNero and Spence Green, who met while working on Google Translate in 2011, the company commercialized the research prototype into a production translation product. Lilt raised seed funding to support expansion of operations. The company intends to use the funds to continue expanding operations and scaling its platform. It is based in Palo Alto, CA.

Led · Equity · Dec 2020

Esports One offers an all-in-one fantasy platform focused initially on the North American and European divisions of League of Legends, combining player research, team creation, and live game viewing. The product relies on proprietary data and analytics plus computer vision that tracks in-game activity and updates player stats without using a publisher’s API. The company says its user base has been growing more than 25% month-over-month. Esports One uses a virtual currency called Bytes for contest entry and customizations, and plans to allow real-money purchases of Bytes starting next year. It is building native iOS and Android apps (currently available via desktop and mobile web) and intends to expand fantasy competitions to Rainbow Six: Siege, Rocket League, Valorant, and Fortnite. The team recently signed a partnership with ESL Gaming. Esports One is a Los Angeles, CA–based data and analytics company for esports fans. Founded in 2017 by Matthew Gunnin (CEO), it uses proprietary computer vision technology to capture live broadcasts in real time and machine learning to interpret them. The company generates real-time information to display to fans, who have control over which metrics they see during each game. Available metrics include player history against opponents, farming efficiency, player positional tracking, previous head-to-head results, objective tracking, and prediction models. Esports One closed a $3M seed funding round co-led by XSeed Capital and Eniac Ventures, with participation from Crest Capital. The company intends to use the funds to continue to scale operations and hire talent.

Participated · Series A · Aug 2020

Narrative operates a data marketplace and software for managing data transactions and has launched the Data Streams Marketplace to simplify buying and selling data. The new product applies e-commerce and search models so buyers can discover datasets, add them to a shopping cart, subscribe or pay by credit card, accept terms, and check out. Data on the platform is sold by third-party sellers who are vetted by Narrative; buyers can learn about sellers and potentially establish direct relationships to assess compliance and quality. Narrative says it will not universally grade data quality or guarantee regulatory safety, but aims to provide transparency and is building a marketplace of third-party applications that could include data-quality scoring tools. Financially, Narrative raised $8.5 million in a Series A and had previously raised $5.3 million, according to Crunchbase. The company plans to use the new funding to hire in product, engineering, sales, and marketing and has emphasized hiring across North America and outreach to underrepresented groups. Narrative operates a data marketplace and provides software that helps companies buy, sell and manage data transactions beyond the marketplace. Its platform addresses fragmentation in formats and transaction workflows, offering reporting and monitoring tools that many firms historically built in-house. Founder and CEO Nick Jordan says customers use Narrative to perform due diligence and assess GDPR compliance on acquired data. The company saw a drop in data availability in the EU after GDPR, but also reported stronger customer response in Q2–Q3 following the regulation. Narrative positions its product to give full transparency between buyers and sellers and to grow customers’ data businesses. The company plans to use new funding to double down on sales and marketing to demonstrate product–market fit and pursue a larger round if successful. Narrative operates a marketplace where data buyers can browse sellers, bid on data, and sellers can manage transactions. Founder and CEO Nick Jordan, an adtech executive formerly at Tapad, built the company to reduce the cumbersome, one-off nature of data deals. The startup describes its approach as “data commercialization” and distinguishes itself from data management platforms. Its initial focus is adtech, though the team says the platform could be useful in other verticals (for example, commercial real estate). Narrative emphasizes privacy and compliance, performing a “pretty thorough review” of sellers to ensure legal and best-practice adherence. Financially, the company has raised $2.25 million across two seed rounds.

Participated · Series A · Mar 2020

unitQ provides an AI-powered customer feedback platform that helps product builders, engineers, support leaders and team members understand feedback in real time. Its platform lets clients discover quality issues, identify which product launches, releases or evergreen features are causing the most bugs or support tickets, and drill into root causes by source, platform, device and customer segment. Led by CEO Christian Wiklund, the company helps teams build products, fix bugs and resolve support issues at scale. unitQ is based in Burlingame, California. The company received an investment from Zendesk Ventures; the amount was not disclosed. unitQ says it will use the funds to expand operations and accelerate development efforts. According to the CEO, the partnership with Zendesk also provides access to a community, industry networks and tailored mentorship to support growth. unitQ uses AI to aggregate and analyze user feedback from public sources (app reviews, social media) and private sources (support tickets, chats, surveys) to surface actionable product-quality issues. Its platform integrates with 26 platforms and ingests feedback via an API to automatically tag and analyze quality problems for engineering, support, product ops and product management teams. The company is primarily focused on consumer companies but also serves some B2B clients; customers include Chime, Pandora, The RealReal, NerdWallet, Strava and AppLovin. unitQ says customers on average increase product quality by 20% within 30 days, and CEO Christian Wiklund reports the company has been tripling its ARR every 12 months. Founded in 2018 and based in Burlingame, California, unitQ plans to use new capital to beef up its engineering and go-to-market teams. unitQ provides unitQ Monitor, a platform that ingests omni-channel user feedback in any language to identify, quantify and prioritize product bugs. The solution surfaces indisputable data to guide engineering and product teams on what to fix next, aiming to increase user retention, decrease support ticket volumes, and align organizations on priorities. The company was founded in 2018 by ex-Skout founders Christian Wiklund, David Eklov and Niklas Lindstrom and is based in San Mateo, California. During a short beta, unitQ identified 3,500 unique bugs in 5.6M pieces of customer feedback data. On average, customers reported 20% fewer bugs within 30 days of adopting unitQ. The company has launched out of beta and is scaling its commercial efforts.