Prompting the Future: Synthetic Media Transforms Creative and Enterprise Outputs
Elizabeth Wu, MBA candidate at Northwestern Kellogg School of Management, Tensility Intern
Armando Pauker, Tensility Venture Partners
Wayne Boulais, Tensility Venture Partners
Synthetic media refers to any type of media – such as images, videos, voice/audio, and text – that is generated by artificial intelligence (AI). AI-driven media employs algorithms trained on large datasets of existing media to learn and generate new content that can be near indistinguishable from human-created content. Synthetic media can provide two outputs: creative and knowledge-based content.
Before the advent of synthetic media, the process of producing media was initiated by intention and created by humans with tools, which brought about a final product.
Intention is the commitment to create a certain output. For the creative process, the source of the intention could be imagination, life experiences, or inspiration. For the knowledge-based content, the source of the intention could be a business goal, data, a medical procedure, or customer support.
Now, with synthetic media, humans and conventional tools are replaced by AI.
One unique aspect of synthetic media is the use of prompting to communicate human intention to the AI model. Certain parameters need to be defined for the model to provide a useful output.
These parameters include commands (what needs to be done), context (how), and task
description (what the output should look like). Prompting is also unique in that it is done in
natural language, not in a stylized code (like a SQL query). Lastly, prompting does not
necessarily require that the user understand the details and intricacies of the tool being used,
i.e, the AI model.
Creative versus Knowledge-Based Outputs
For simplicity, we will categorize media into two types: either creative or knowledge-based.
Creative media examples include images, marketing/ad copy, or videos, which were traditionally created with tools such as Photoshop, Word, or Premiere Pro, respectively. Knowledge-based content, especially that pertaining to business processes, includes spreadsheets, presentations, or documents, now produced with Excel, PowerPoint, or Word, respectively. This type of content could transform data to images or ideas, for example, when numbers in Excel are collected and displayed as a visual graph that would in turn be embedded into PowerPoint.
With synthetic media, AI engines and prompting replace human-driven tools to produce a given output, but presumably with higher productivity. In creative processes, new AI tools such as DALL-E (images), ChatGPT (ad copy), and Synthesia (video presentations) can produce the same media as traditional applications that are near-identical to human-generated content. But in business, not many prompt-driven AI tools currently exist that meet enterprise output standards for software applications. However, Videate is the exception for customer training and product support videos.
Synthetic Media in Enterprise Video Production for Customer Support
The customer support benefits of videos are evident when we consider that most people do not want to read long product manuals or instruction guides because specific information is hard and time-consuming to find. Most people are relegated to searching YouTube or TikTok for videos that may be self-produced or outdated. Companies should make these videos, but they have been a heavy lift to produce.
One company, Videate, has created automated video production tools that give companies the ability to scale video production without manual video editing and human voice recording. Videate specializes in explainer videos for software content that automatically refreshes every time a new product version or UI change is released. This improves the quality of onboarding and decreases the “time to value” for a software user, who is able to learn more effectively and find answers to questions and troubleshooting issues without having to wait on support.
Consistent with other generative AI tools, Videate expects that eventually users will be able to use prompting to receive an AI-generated video for an up-to-date answer to any software
product question. There would be no lag time between the need and the video answer.
The Need for Human Verification
While synthetic media can enable creative and knowledge-based enterprise processes, it is not perfect. Synthetic media can be subject to errors and biases, just like any other form of AI.
Hallucination issues with generative AI have been well documented and are slowly being
understood. When used in any type of content creation workflow, we must continue to
incorporate human overseers for quality assurance (QA) to catch errors and ensure the content
is factually accurate and meets all legal and ethical standards. Additionally, human QA can
prevent the misuse of synthetic media, such as creating misinformation or harmful content.
By combining the power of synthetic media with human oversight, everyone can achieve greater efficiency and productivity while still ensuring that their content is accurate and trustworthy.
A Dual Approach to AI Product DEvelopment: Agile and Periodic Enhancement cycles
A collaboration between Tensility Venture Partners and Actuate.ai.
Wayne Boulais and Armando Pauker Managing Directors at Tensility Venture Partners
Brian Leary VP of Product & Operations at Actuate.ai.
Software development teams have embraced Agile methodologies for continuous product enhancements to deliver working software rapidly and to ensure constant improvement for over 20 years. Hardware development teams embrace periodic product enhancement with longer development cycles focused on the introduction of major new features and hardware cost reductions. AI teams need both approaches. This blog concludes with a case study at Actuate AI, a NYC computer vision AI company focused on physical security.
AI Development Requires New Processes
AI product development teams must think differently and holistically about their software development and dev ops processes. On the one hand, AI teams need to maintain the benefits of agile development noted above for improving models or introducing new capabilities. On the other hand, these teams are also tasked with the cost reduction of cloud compute and storage and may periodically switch models in production as better ones are available. This requires two different development methods concurrently: Agile methods for continuous monitoring and enhancement of models in production and a periodic product enhancement cycle for sustainable AI products. These two methods are required whether a given model is maintained and trained in-house, or whether it is accessed through APIs (like many new Large Language Models) and fine-tuned locally.
Agile Methods for AI Product Development
Agile development involves a series of scrum sprints, typically two weeks in length, to deploy code continuously. This continuous product enhancement process is designed to reduce the complexity, cost, and disruption of creating and introducing new software into production. Agile methods enable AI development teams to closely monitor and enhance models in production. AI models are constantly tuned and retrained to correct false positives and false negatives, to extend the features in the model, to correct issues of bias and drift, and to respond to new data sets or trends. The Agile approach ensures that AI products remain up-to-date, accurate, and relevant in rapidly changing environments and reduces the risk of negatively impacting production processes.
The Need for a Periodic Product Enhancement Cycle
In addition to the Agile methodology, AI products require a periodic product enhancement cycle. This cycle consists of periodic systemic changes due adjustments in product usage, new performance needs, model upgrades and reduced operating costs.
For example, these longer redesign cycles allow AI teams to evaluate potential infrastructure changes that can lower operating costs while maintaining or improving AI engine performance. Operating costs could be lowered through numerous ways such as changing cloud configurations or vendors or making internal model tradeoffs to reduce retraining compute costs. These improvements can require more analysis and testing than a normal two week sprint.
This approach is similar to how hardware engineering teams develop upgrades. The AI architecture — consisting of storage, data, training, and algorithms — is embedded in the cloud, and the enhancement process requires assessing various attributes, including: storage/IO options, model version, algorithm performance, API calls, data ingest and cloud computation costs before making a significant, systemic change.
The product enhancement cycle is an architecture upgrade that is similar to the months required to introduce a new hardware revision into production compared to software sprints. An example of a complex system that follows a similar hardware upgrade cycle is the iPhone, which typically has a production upgrade cycle of about 12 months.
Experience from Actuate’s AI product development team
Actuate offers real-time AI video analytics for surveillance systems to detect threats to safety and security across a broad range of use cases and commercially available hardware. The product does not require any hardware installation. All of the compute resources are in the cloud, from video ingestions and processing to AI model training and deployment.
The focus of Actuate’s AI product development team is to provide a stable, scalable, and cost-efficient architecture where the models can be constantly retrained and deployed with minimal intervention. Both software engineers and data scientists work closely together during the continuous upgrade cycles of AI model improvement and major architectural upgrades. Actuate operates 7–10 centralized AI models that operate independently and are retrained continuously from data received from customers.
The team quickly learned that each customers’ usage is not stable throughout a day, week, or even a month. More efficient resource scaling was needed to accommodate the variability while improving margins. Moving to a containerized system through Amazon ECS was a major overhaul that allowed for horizontally scalable resources to be spun up and scaled on demand. Horizontal scalability guaranteed resource allocation for each customer, but it did not provide cost effective resources at scale, especially given the changes in throughput of our customers daily. Following the transition to Amazon ECS, the team then identified the benefits of Kubernetes, the container orchestration tool, to allow for even better resource utilization (vertical scalability with larger resource clusters) and a higher level of automation for the retraining pipeline. The goal during each upgrade was to keep our customers running with no interruption or indication that the backend change was happening. We followed the process below, combining product enhancement cycles and agile methodologies:
Repeat steps 4–6 until 100 percent
Steps 1–4 require a longer, product enhancement cycle similar to a hardware build. Steps 4–6 are then repeated following an agile methodology as the entire cycle can be completed in 1–2 weeks, ending when everything is successfully transitioned.
The ECS migration addressed the high cost stream ingestion layer, but the model development process was still in need of an enhancement. Prior to any automation pipeline being built, lengthy, manual retraining was bogging down our team. SageMaker provided the building blocks required to automate most of the training process, with the added benefit of automated deployment mechanisms that can be built to link the operating and development environments. Much like the architectural upgrades above, the SageMaker implementation followed a similar, months-long process, culminating in full integration after the transition to Kubernetes. The results are short model improvement sprints (1 week or less) where multiple models trained are evaluated and tested simultaneously, with the most applicable being released instantly.
AI product development teams can benefit from a dual approach that combines Agile methods for continuous model monitoring and enhancements with a periodic product upgrade cycle for sustainable AI products. This combination ensures that AI products remain up-to-date, accurate, and cost-effective while delivering the best possible performance to users. As the AI landscape continues to evolve, embracing these approaches will be crucial for creating AI products that thrive in the market.
Web3 Adoption in Healthcare
Wayne Boulais, Tensility Venture Partners
Armando Pauker, Tensility Venture Partners
Nicholas Shapiro, UPMC Enterprises
Vicky Wang, Kellogg School of Management, Northwestern University
There are categories of healthcare problems that are well-suited for Web3 infrastructure solutions. Existing healthcare institutions are limited by practices and regulations that do not incentivize data sharing across participants. Key issues in healthcare data today include the lack of a full longitudinal data set for a given patient and the struggle by patients to control their own data. With Web3, we can rethink how the participants in the ecosystem can be incentivized and energized to find new ways to share data to improve the healthcare system. Web3 concepts suggest new approaches to addressing the issues of trust, alignment, and transparency.
The Decentralized Autonomous Organization (DAO) is a flexible organizational structure that could be the first way that Web3 ideas begin to impact healthcare. The Web3 smart contract and the blockchain provide the means for incentivizing data sharing in many forms across the healthcare landscape, which consists of patients, physicians, healthcare providers, payers, pharmaceutical companies, and medical research institutions.
The graphic below is a general model of a healthcare DAO. At the highest layer, the founding team determines their mission and strategy. The founding team, in discussion with other participants, determines how data is going to be shared (with some defined opt-in procedure to meet privacy regulations), the incentives for sharing the data, the group governance structure, and how the treasury will be managed. The last layer encompasses the potential constituents of the DAO. Every DAO will have a different set of constituents based on the mission.
Examples of existing and potential healthcare DAOs include:
Of course, there will be distributed healthcare communities that do not use the DAO structure. For example, Hashed Health has started a company, ProCredEx, that “allows members to sell or share verifications they’ve created and purchase verifications they need. ProCredEx’s validation engine and distributed ledger technology securely presents immutable credentials data.” The benefits to the health system are participating in a federated sharing of credentials across all its employees - not just the physicians - to lower the cost for all providers.
Importantly, not all healthcare problems are best addressed by Web3 infrastructure. Careful consideration must be given to matching the appropriate enabling technology with the proposed solution. Deploying Web3 infrastructure for the sake of describing a company as forward-looking will simply add cost and friction to stakeholder adoption. Several examples listed above demonstrate areas where the information sharing that does take place is controlled by centralized clearing houses with significant market capture. Web3 infrastructure, when directed effectively, can erode the control of middlemen that extract meaningful value from stakeholders while providing limited improvement - which leads to a lack of trust and limited sharing. Healthcare innovation using Web3 infrastructure will ultimately succeed in use cases where data creators and owners are empowered and the whole is materially more valuable than the sum of the parts.
*UPMC Enterprises, a division of health system UPMC, owns a financial stake in Hashed Health.
We are excited to announce our investment in Culina Health’s seed round. Based in New York City, Culina Health is a clinical telenutrition platform focused on treating, reversing, and preventing chronic conditions using medical nutrition therapy from registered dietitians (RDs). It provides data-powered personalized nutritional programs, enabled by a patient-facing app, for patients, allows for seamless electronic health record data transfer and feedback for physicians, and provides back-office and administrative help for RDs.
Clinical nutrition centers on the prevention, diagnosis, and management of nutritional changes in patients linked to chronic diseases and conditions. While the wellness nutrition space already sees an abundance of technological solutions such as nutrition apps, clinical nutrition is still highly reliant on an individual RD-based approach with limited technological solutions available and no market leaders, despite the market growth. Clinical nutrition support will provide value for a large population with chronic disease (diabetes, celiac disease, cardiovascular disease, etc.) and other nutrition support needs. There is a huge opportunity in bringing an effective solution to this space.
Culina has developed “The Culina” way of care, where it provides extensive training and education to the RD staff to ensure standardized and quality care. In addition, care delivery data models are developed from data captured from the platform and smartphone, further facilitated by AI-enabled features such as photo interpolation of food to-be-consumed.
We are excited for the founding team to revolutionize the clinical nutrition space. Vanessa Rissetto, RD, co-founder and co-CEO, has 10+ years of experience as an RD and was a dietary director at NYU. She is a highly sought-after dietitian with frequent appearances in the press and media. Steve Kuyan, co-founder and co-CEO, has 10+ years of experience in startup, venture capital, investing, and startup growth. He founded and led NYU Future Labs with 40 exits. Tamar Samuels, RD, co-founder, has 7+ years of experience in hospital and private RD practice. Given the management team’s expertise and recognition in the clinical nutrition space and their experience in start-up operations and growth, Culina Health is uniquely positioned to attract top RD talents and provide quality care to patients.
We believe in the founding team’s vision, and we are excited to partner with them to achieve their mission: democratizing access to clinical nutrition to increase everyone’s healthy life span through personalized data-driven telenutrition!
Web2 to Web3 Convergence in the Enterprise
Wayne Boulais, Managing Director, Tensility Venture Partners
Armando Pauker, Managing Director, Tensility Venture Partners
Alex Poon, Co-Founder, CharmVerse
Distributed Autonomous Organizations (DAOs) became possible with the advent of blockchain and Web3 infrastructure. DAOs are appealing because of decentralization, community building and the promise of shared outcomes. The participants in a DAO can have more input on the direction and governance of the organization than the average employee in a traditional corporation. The flexible structure allows individuals across the globe to seamlessly come together and build a community centered around a common objective. And the token incentive structure allows all the members of the organization to benefit from the collective success of the project.
The first DAO, The DAO, was launched in 2016 on the Ethereum blockchain as an investor-directed investment fund that crowd sourced funds with a token sale. This DAO quickly ran into problems when its smart contract was hacked allowing the hacker to drain 40% of The DAO’s treasury. At the time of the hack, the Ethereum blockchain was only about one year old. In a controversial move, the Ethereum blockchain instituted a hard fork that rolled back the Ethereum history to before the hack. This change allowed the stolen funds to be returned to investors. (See What was the DAO?)
Since 2016 the blockchain ecosystem has gained acceptance and credibility in crypto trading, decentralized finance and non-fungible tokens (NFTs) for collectors and creators. This acceptance has led to renewed interest in DAOs as an organizational model and to fast growth in the number of DAOs. According to DeepDAO.io, in March of 2022, more than 600 new DAOs were formed, and they are now tracking nearly 4,900 DAOs. DAOs are growing at more than 160% annually. Yury Lifshits, the founder of SuperDAO.co, has predicted that one million DAOs will exist in the near term.
The ability to organize and manage people, tasks and money in a decentralized organization is a challenge. Blockchains, smart contracts and tokens are the critical pieces of enabling infrastructure that allow organizations to create and align incentives in a decentralized fashion. Crypto wallets are an essential control point for the DAO to manage role-based access, voting rights and payment methods.
Numerous DAO tooling companies have been created recently to assist in the formation and management of these decentralized organizations (see market map below).
Source: Nichanan Kesonpat, Platform & Content @1kxnetwrk, @nichanank
While there are many different DAOs forming, there are four broad categories of DAOs.
Investment DAOs gather funds to make shared investment decisions in companies, grant distributions or shared collections of art, NFTs and other collectibles. In 2021, the Constitution DAO raised over $42M attempting to buy a copy of the US Constitution. Links DAO is intent on buying a golf course in 2022, while Krause House DAO is a collective effort to buy an NBA team. These DAOs have a strong focus on the protocol they are built upon because a DAO can have large, shared treasuries that require security, transparency and smart contracts to distribute funds into projects, pay contributors and document the approval process.
Social DAOs bring a community together to collaborate around shared interests. They focus on communication methods, community content, and shared values using a token gating process. For example, the Friends with Benefits DAO is a group of artists, thinkers, and Web3 enthusiasts attempting to bridge culture and technology with a creative, Web3-native ethos. The $FWB token enables governance and ownership for the members that share the DAO’s collective values.
Entertainment DAOs enable community driven content decisions for people like film makers, authors, artists and other creatives. These DAOs allow creators to leverage their brand to build a broader ecosystem of content delivered by community members across various channels (podcasts, newsletters, social, etc.).
Product and Services DAOs provide various business offerings. For example, RaidGuild is a design and development agency for the Ethereum ecosystem. Juicebox is a platform for fundraising from the community and building a treasury. DAO Masters and DAO Central are showcasing DAO projects.
Key elements of a DAO:
It is early days for DAOs, but things are moving very rapidly. As DAOs grow and demonstrate success in the Web3 eco-system, we see the potential of their novel organization techniques, tools and processes to be adopted initially where Web 2 to Web 3 businesses converge. Some potential examples are:
Here is a thoughtful blog post by Mitch Worsey, our intern at Tensility Venture Partners, discussing Web3 dynamics and how that may affect Media and Entertainment.
WEB3 PATH TO ENTERPRISE ADOPTION
Wayne Boulais, Co-Founder & Managing Director, Tensility Venture Partners
Armando Pauker, Co-Founder & Managing Director, Tensility Venture Partners
Paul Hsu, Founder and CEO, Decasonic
We, the authors, had the pleasure of connecting our respective areas of investing and operating experiences across enterprise and consumer platforms. Our collaboration takes a fresh perspective at Web3 enterprise adoption. We are energized when sets of expertise come together to innovate a frontier technology.
In this blog post, we are considering what conditions need to be met for Web3 technologies to be ready for enterprise adoption. Web3 encompasses open and decentralized networks, built on blockchain layers, and that employ some form of tokens as a mechanism for collaboration and alignment of incentives. We see this as separate from the current discussions on the metaverse, which encompasses AR/VR components to enhance the end user Internet experiences.
Traditional enterprises will embrace Web3 when they see a path to new revenue growth opportunities: new products, new business models, new customer segments, new geographies, new sales models, and / or new workforce structures. New business models are historically introduced first by disruptive startups and embraced by enterprises later, as the technology matures. Web3-focused hackathons and accelerators today are the places where new ideas are born and incubated. These hackathons, many sponsored by specific blockchains, drive collaboration and creativity with blockchain native organizations. These groups are driven by ad hoc technical teams which create "projects" or applications that can become companies.
The first successful use cases in new technical segments are often focused on end-consumer applications that lead to widespread adoption. This echoes the trajectory of adoption for AI. The first major adoption for AI in the corporate realm was in companies like Facebook, Amazon, Netflix, and Google that collected massive amounts of user data and used that to drive recommendation engines and increase the engagement of each company’s particular content (e.g., ads for Google or movies for Netflix). There are several examples of the early consumer applications of Web3.
1) Bitcoin was consumer-driven for a decade led first by technical early innovators, then retail investors before the leading edge trading firms and hedge funds. Now some e-commerce businesses, some corporate 500 treasuries, and a few governments have moved forward with limited adoption.
2) The NFT (non-fungible token) technology burst on the scene first through the EIP 721 standard in early 2018 and then scaled towards mainstream awareness in 2021: first, video moments with the NBA, then art collectibles, and now sports and celebrity influencers seeking new ways to engage with fans and followers. The blockchain provides clear provenance of ownership, a critical value driver for digital products or services. More recently, AMC and other retail brands began initial engagements - as part of loyalty programs through limited minting.
3) P2E (Play-to-Earn) business models have been introduced for playing online blockchain-based games where NFTs are minted for gameplay and players earn digital coins or tokens. Dramatic experimentation with business models including staking NFTs in games to win more gameplay tokens and the use or sale of NFTs across games is in process now.
4) Staking, verifying transactions, renting, or lending of digital coins for specific periods of time in order to collect fees in the form of more coins is a financial innovation using the peculiar aspects of decentralization and blockchains. Today we see the rise of DeFi (decentralized finance which allows individuals to engage in practices known as yield farming.
5) The DAO (distributed autonomous organization) concept has initial consumer traction in several forms. Bankless DAO started from a podcast (“Bankless”), but has become a sprawling network of new businesses/services created by and for the community. Other examples are the ConstitutionDAO (and its failure to outbid Ken Griffin for the US Constitution) or LinksDAO (with its goal to acquire a golf course exclusive for members).
The infrastructure for these consumer applications is maturing quickly. Crypto exchanges (Coinbase, FTX, Binance) are accessible to the mainstream, simplify the buying and trading for fiat money for digital coin, and enable the purchase and sale of NFTs. Trades are allowed between different digital coins. These exchanges act as central points of liquidity for digital native transactions. Marketplaces like OpenSea have arisen for consumers to efficiently search, find, buy and trade tokens.
For enterprise adoption of Web3, we consider several conditions. There was a previous effort in 2017 to use the blockchain for enterprise applications, but that adoption did not occur. A perspective on why this is different now will be discussed in a subsequent blog post.
1) Scalable infrastructure must be available from the point of view of both cost, availability and speed. The cost of using We 3 resources must approach the cost of cloud hosting today. In Web3 there is already a migration to proof of stake for Ethereum to address the high transaction costs caused by the proof of work approach in Bitcoin and Ethereum 1.0. New alternatives, such as Solana, Polygon, Avalanche, Near and Wax, have been purpose-built to address the cost issue. Other innovations in Proof of Work include Kadena’s chainweb blockchain architecture.
The transaction speed must be capable of communication throughputs similar to current internet and transaction processing speeds. Already new blockchains, like Solana and Avalanche, have been engineered for high transaction speeds. Solana claims to be capable of 50,000 TPS with the combination of proof of stake and proof of history approach. This limit approaches Visa's TPS.
2) Identity/ privacy/ security infrastructure must be accepted, however, the ideas of privacy and identity are being redefined in Web3. The digital wallets used to store digital coins and NFTs have curious properties and implications in this area. These wallets are anonymous, but the contents are visible because the transactions leading to the wallet contents are immortalized in the open public ledger or blockchain. Wallets become the de facto identity in the Web3 world. However, since you can have multiple wallets, you can have and curate multiple identities. This is unique to Web3 and allows for individuals to segment reputations through the contents of their wallets. The wallets can prove, for example, that the holder has NFTs that were only given to attendees of certain technical conferences or to show technical literacy through the earning of badges. These NFTs act as validations of activity or ownership.
3) An understanding of tokens and their power to re-align incentives. Part of the allure of Web3 is the aligned user and growth incentives afforded by the use of tokens. These tokens are the means by which identity or reputation can be established for the purposes of access control, collaboration, or incentive structures. The potential for the increase in value of the tokens also sets the stage for a change in the software licensing model. The transfer of a token as part of a software licensing agreement would allow the buyer of software to have a stake in the success of the software developer and a say in the community of customers that use the software. The more the business model relies on key elements being on the blockchain, then the more likely a token will be a good incentive and governance addition to the business model. Conversely, non-blockchain companies will be very challenged in using tokens because they are not likely to be cost-effective, value-add, or strategic.
4) New possibilities for customer segments or employee organizations. The ability to see inside wallets while keeping identities private brings a new world of possibilities in customer segmentation and data control for the end-user. The ability for the end consumer to control and curate multiple identities may allow for the end consumer to monetize data (through the staking of a specific wallet) that is now gathered, controlled by, and profited by large data economy tech companies. DAOs could allow users to be part of entities where the benefits of unique membership, through special access or unique value, are validated by tokens. We see the potential for DAOs to become enterprise customer advisory groups or select corporate influencers or interest groups approved by HR in distributed organizations once the new governance models become more familiar to businesses.
Like many emerging technology adoption, the Web3 path to the enterprise will be bumpy with many starts and stops. Enterprises will eventually adopt and change after disruptive startups show the way to new customers and revenue, in much the same way that Google and Amazon did in their formative time decades ago. We are excited about this future.
Tackling the Turing Test: Automation Quickly Creates Video Content for Customer Success
Customers of B2B SaaS products depend on videos as their preferred vehicle for customer success information. Many companies maintain vast video libraries that offer a wealth of narrated information in easy-to-follow screen recordings, and customers often prefer this more dynamic format to written instructions. Unfortunately for the enterprises, keeping this library up-to-date is a highly manual process: video recordings take time to produce, are costly to translate into other languages, and – most plaguing of all – can become quickly outdated with subsequent product release cycles. It is not unusual for companies to accumulate ‘video debt’, or a glut of outdated videos that do not accurately represent the latest product version.
We at Tensility understand the importance of Customer Success for B2B SaaS companies, and are excited to announce our investment in Videate’s seed round. Based in Austin, Texas, Videate’s technology uses AI to quickly transform written documentation into finished videos related to support, on-boarding, or training. The manual tasks of recording, adding voiceover, and editing video content are completely automated with an easy-to-use platform that leverages browser automation, NLP, and text-to-speech technology. Videate’s AI eliminates the hassle of software video creation, offering special effect solutions and pronunciation tools to ensure the videos look, feel, and sound like they are human-made. Videate's automated language translations can transform a single piece of content into a globally-relevant resource. With an initial focus on B2B SaaS companies, we are very impressed with the team’s early traction and customer feedback that likens the previously burdensome process to magic.
The management team is made up of a seasoned cohort of leaders and innovators. Co-founders Dave Gullo and Mark Hellinger are former C-suite executives with decades of experience in online video and enterprise application construction. They are highly motivated to tackle challenging problems, and their early traction speaks to their success at addressing a common pain point for SaaS companies, product managers, and B2B customers. We are thrilled by their product roadmap that demonstrates a deep understanding of how to apply AI to delight customers. We look forward to partnering with Videate to transform Customer Success and help B2B companies meet their customers’ needs with ease!
HARNESSING AI IN QUANTUM COMPUTING
We are excited to announce our investment in Agnostiq’s seed round. Based in Toronto, ON, Agnostiq Labs (“Agnostiq”) is developing a suite of applications that seamlessly enable financial institutions (FIs) to leverage QC to power their trading algorithms, all without the need of expensive quantum programming resources. Agnostiq’s applications will allow banks to securely transpile, optimize, and deploy quantum machine learning (QML), quantum Monte Carlo and quantum neural networks in portfolio optimization models. This breakthrough will lead to development of new algorithms at a faster rate and better market predictions.
Financial institutions (FIs) have realized that classical computing is approaching the limit of Moore’s law and there is a gap between the amount of new unstructured data coming in and their ability to process that data to make better decisions. As such, FIs and other organizations are looking to quantum computing (QC) to resolve these issues and deliver the high performance computing they need to maintain their edge. Whether developing the better algorithm to price options or conducting trades in one 64 millionth of a second to capitalize on the arbitrage opportunity, FIs need the compute power to analyze hundreds of millions of data points in real-time to execute their strategies.
While initially focused on banks, we see tremendous opportunity to apply their unique quantum AI workflows and best-in-class security applications across several industries, such as pharmaceuticals, healthcare, and energy. Additionally, we believe Agnostiq can become the platform-of-choice that accelerates enterprise adoption of QC and transforms data scientists into quantum scientists.
We are incredibly impressed with the management team and talent. Collectively, the founders Oktay Goktas, PhD, and Elliott MacGowan possess the unique combination of deep technical knowledge in QC, sales acumen and operational expertise needed to succeed in this nascent industry. Agnostiq has also proven its ability to attract top tier QC talent with backgrounds in quantum physics, cryptology, mathematics, and computational physics. Given the dearth of available and qualified talent, Agnostiq’s ability to attract, hire, and retain top scientists will solidify its market position and competitive moat.
We believe in Oktay, Elliott, and Edwin’s collective vision on quantum computing and are excited to partner with them in their quest to help enterprises harness the power of AI in quantum computing!
Today companies are driving workforce productivity using 3rd party SaaS applications like Salesforce, Jira, Dropbox which form the backbone of critical functions across the organization. According to a 2020 Devsquad study, companies deploy, on average, 34 different SaaS tools, and that number increases dramatically with company size (reference 1 below).
Company tech stacks are also becoming increasingly decentralized and thus harder to manage. Communication tools such as Slack and Microsoft Teams have made it easier than ever to collaborate across teams, but have also led to sensitive information moving unchecked across the enterprise, increasing the risk of data loss and leakage. Last year, over 165 million records were either lost or exposed within the US alone (reference 2 below), all the while increased regulation (e.g., GDPR and the California Consumer Privacy Act) has levied significant fines and put additional pressure on businesses to revamp their information security and data loss prevention (DLP) strategies.
As organizations continue to grapple with this changing landscape, we are thrilled to announce our investment in Polymer Solutions’ (“Polymer”) seed round. Cyber security is one of our key investment areas, and continues to be important as more work moves to the cloud. We are proud to support Polymer’s vision to become the preeminent platform that manages DLP and redaction across the enterprise technology stack. While most of the DLP market has focused on encryption-based solutions for at-rest data, Polymer is unique in its approach by applying Natural Language Processing (NLP) on unstructured, in-motion data to redact Personal Identifiable Information (PII) and sensitive corporate information. This solution will be critical to companies that collect sensitive customer information, such as financial or health data.
Polymer provides an easy-to-use solution that allows enterprise Information Technology departments to create access controls as well as monitor, secure and redact sensitive data across dozens of collaboration apps, including Slack, Github, Dropbox and Zapier. Polymer’s product has already gained significant traction with several large enterprise clients.
Polymer was founded by Yasir Ali and Usman Malik, who are supported by a strong technical team of cyber security, big data and machine learning experts who are excited and well-positioned to tackle this complex data redaction problem. We are thrilled to partner and support Yasir, Usman, and the Polymer team as they continue to expand their data governance platform and grow into a leading security company!
1 ”60 SaaS Statistics and Trends for 2020”, Industry study via Devsquad, 2020
2 “Annual number of data breaches and exposed records in the United States from 2005 to 2019,” Identity Theft Resource Center, 2019
Armando and Wayne