July 2022 Summaries

This book series separates facts from fiction in the startup journey, providing guidance on hard-and-fast rules to increase chances of survival. The Symbl.ai for Startups team recommends 10 books that will help founders start their journey as successful entrepreneurs off on the right foot. These books cover various aspects of starting and running a business, including leadership, customer service, design, venture capital, lean startup methodologies, networking, and more. They offer actionable advice, insights into human-centered design, and strategies for building valuable businesses that offer something entirely new to the world.

Symbl.ai has achieved SOC 2 Type II certification, demonstrating its commitment to delivering a highly available and secure conversation intelligence service. This certification ensures that the company's systems, controls, and processes meet the cybersecurity standards outlined by the American Institute of CPAs, providing assurance to customers about the protection of their data. The SOC 2 Type II report provides detailed information on Symbl.ai's security policies, encryption protocols, access controls, change management process, data backup and disaster recovery strategies, system monitoring, and more. This certification is a non-negotiable aspect of what Symbl.ai does, and it provides additional assurance that the company has the right controls in place to protect customer data and ensure service availability.

Conversation intelligence is a process of analyzing conversations in text, audio, or video files to understand their meaning and inform decision-making. It can be applied to various channels such as customer service exchanges, sales interactions, marketing efforts, e-commerce transactions, and customer experience. The technology must fit the problem and provide valuable analytics beyond content, including talk time and speaker ratio, silence, pace, overlap in conversation, entities, and intent. By leveraging conversation intelligence, businesses can improve their processes, retain customers, increase revenue, and enhance the overall customer experience. A tool like Symbl.ai provides a platform that connects through APIs or SDKs to capture conversations, determine features, and provide analytics, making it easy for developers to embed conversation intelligence into applications and boost business.

The integration allows users to ingest or consume Zoom Cloud Recordings before sending data into Symbl for processing, aiming to provide meaningful insights such as topics, questions, action items, follow-ups, and summaries. The system architecture is based on serverless components, automated using the AWS Serverless Application Model (SAM), which provides a guided deployment process. The integration consists of several components, including API Gateway, Lambda functions, S3, SQS, DynamoDB, and FFmpeg, working together to merge audio recordings, upload them to S3, persist metadata on DynamoDB, initiate an SQS message for further processing, and submit the audio recordings to Symbl using the Async Audio mechanism. Users can configure the custom Zoom app webhook endpoint, deploy the application using SAM deploy, and run the application by initiating a Zoom meeting, adding participants, recording, and ending the meeting. The integration also includes best practices for handling limitations such as cloud recording limits, lambda execution time, and Gmail authentication errors.

For the second year in a row, Symbl.ai has been named one of the Top 100 Early-Stage Companies to Work for in 2022 by executive search firm Will Reed, out of 500+ companies that raised Seed & Series A rounds over the past year. The recognition is based on mission, culture, growth trajectory and founding leadership, highlighting Symbl's commitment to building a human-first culture that offers meaningful work and supports the full employee experience. Symbl.ai has recently expanded its engineering and leadership teams with a $17 million Series A funding round, and CEO Surbhi Rathore attributes this recognition to investing in her team and building a company that shares a similar vision and focus. The award is seen as a testament to Symbl's growth plans and its continued focus on investing in its people.

Founders who start their journey by doing what they're most passionate about may shift towards employee management as their startup scales, but it's essential to carve out time for activities that bring inner satisfaction and achievement, such as the flow state of being. This state is characterized by a pleasant engagement with adversity, fostering unselfconsciousness and deep satisfaction. To access the flow state, founders need to eliminate distractions, balance challenge with skill, and push past negative mental chatter. By doing so, they can become better entrepreneurs and stay motivated amid the hecticness of running an early-stage startup.

Integrating Zoom APIs with AWS Lambda offers new possibilities for analyzing meeting conversation data, topics discussed, sentiments, and live captioning. With flexible APIs, developers can analyze and gain meaningful insights from their Zoom data, while events subscription ensures application sync with changes related to resources. However, common challenges arise when integrating with Zoom APIs, such as cross-origin resource sharing issues, complex debugging experiences, handling complex data analysis limitations, and steps to take for integration. To overcome these challenges, developers must create an AWS account, build a deployment pipeline, generate Zoom API credentials, add webhooks, integrate a monitoring/logging tool, and consider alternatives like Symbl.ai, which provides real-time insights from Zoom calls with no need to set up the Zoom API or provisioning infrastructure. By understanding these steps and best practices, developers can successfully integrate Zoom APIs with AWS Lambda and build scalable applications that make the most out of their integrations.

Today's consumers expect a seamless brand experience from initial engagement to checkout, with continued customer support being crucial. However, companies with small teams often struggle to meet these expectations. Chatbots can offer 24/7 sales and customer support, but they're not perfect and require ongoing reiteration and tuning. To enhance the customer experience, businesses need to track key chatbot metrics such as customer satisfaction, completion rates, fallback FBR, goal completion rate, human takeover rate, total tickets created, missed chats, unrecognized customer queries, average handling time, number of bot sessions initiated, new vs. returning users, blockers, conversation logs, and revenue generated. By monitoring these metrics, businesses can identify areas for improvement, optimize their chatbot content, and refine their conversational AI pipeline to provide a better experience for customers. Ultimately, the goal is to use chatbots as a tool to automate customer services and sales while providing a great experience with your brand.

This blog post guides users in processing multi-channel audio recordings using the Symbl Async Audio URL API. To improve transcript accuracy, parameters such as speaker-separated channel audio processing, custom vocabulary, and phone mode can be added to the request body. The Conversation APIs provide insights on topics, follow-ups, questions, action-items, and trackers, which can be accessed after completing an asynchronous job. The APIs also offer additional features like summarization, entities detection, and sentiment analysis. Users can customize their experience by enabling or disabling various options such as speaker ratio, talk time, and silence analysis. The post concludes with the usage of Trackers for getting additional insights that cater to business needs.

The text discusses extending the scope of chatbots and conversational AI experiences without relying on subject matter experts. It highlights the challenges of adding more intents and conversational flows, bringing context from virtual assistant interactions to live chat and calls, and integrating with different systems of record. The post aims to provide insights into how Symbl can help automate data labeling activities, understand new patterns in real-time, and transfer context between channels. Symbl's platform enables users to process unstructured conversation data, identify unknown patterns or scale the identification of known ones, using machine learning and natural language processing techniques. The text also introduces Trackers, which are user-defined entities that allow tracking critical moments in conversations across multiple use cases in real-time and asynchronously. By leveraging Symbl's Async APIs, users can speed up data labeling, use as a failover intent detection engine, or inform new conversation patterns for the same intent.

Google Cloud Functions, AWS Lambda, and Azure are major serverless services that offer a range of features, pricing options, and developer experiences. Pricing is an important factor, with Google providing twice as many free requests but doubling the cost for additional requests. Features such as APIs, integrations, and support for various languages play a crucial role in choosing the right service. Developer experience is also vital, with Google Cloud Functions aiming to simplify the process, while Azure offers multi-language support and extensive documentation. Reliability and scalability are essential, with all three services offering SLAs and guarantees. Choosing the right provider requires careful consideration of business needs, growth plans, and long-term implications. By selecting the right serverless service, developers can achieve rapid scaling and growth for their AI voice apps.

This content series from Symbl.ai for Startups aims to provide a realistic and actionable guide for early-stage voice AI businesses. The four steps outlined in the article are crucial for success: testing demand for your product, finding spaces where your product belongs, generating content, and making yourself and your startup as visible as possible. Each step requires equal effort and attention from the founder, who must be willing to put in the initial burden of marketing their product, build a landing page, find paying customers, and create a solid digital presence. The article emphasizes the importance of grit, vulnerability, and being honest in one's approach, cautioning against shipping a "piece of crap" too early in search of feedback. By following these steps and persevering through the process, early-stage voice AI businesses can increase their chances of success.

Symbl.ai is an API platform that enables developers to add machine learning to their sales workflows by analyzing voice, video, or text conversations. It provides a range of APIs, including Async API, Experience API, and Summary API, which can be used to process meeting or call recordings, create summaries, and provide insights into customer conversations. The platform's public workspace on Postman allows developers to easily get started with Symbl.ai's APIs and automate sales processes from cold calls to follow-ups. With Symbl.ai, developers can reduce the time it takes to make an API call and access pre-built experiences that simplify the analysis of vast amounts of conversation data. The platform is designed to help developers create new experiences throughout the entire communication touchpoints of the sales process, enabling them to transform recorded calls into intelligence for their sales teams.

REST is a stateless architectural style for building application programming interfaces (APIs) that follows constraints and principles to ensure understanding between applications on the Internet. It's characterized by unidirectional communication, where the server only gives data when requested. In contrast, WebSocket is a two-way communication channel with a server, enabling both sending messages to the server and receiving event-driven messages back. WebSocket is a communications protocol that uses HTTP links, whereas REST is a style of building APIs. While REST is simple and easy to use, WebSocket offers performance gains over REST in certain cases due to its stateful nature. However, using WebSockets for simple applications can complicate the development process. REST is well-suited for isolated, occasional communications, such as creating basic APIs or GET requests, whereas WebSocket is ideal for real-time interaction and sending events back to users. Ultimately, deciding between REST and WebSocket depends on the specific needs of the project, with REST being the most common choice for simple web projects and WebSockets being better suited for applications requiring real-time information exchange.

The Symbl for Startups content series provides guidance on crafting a stellar pitch deck, emphasizing that less is more, eliminating unnecessary information, and keeping the design sleek and minimal. It's essential to understand your audience, tailor your pitch deck accordingly, and continuously update it as your business scales. The ideal structure of a pitch deck includes an intro, problem statement, solution, product, business model, go-to-market plan, competition, team, progress, and contact information. Key mistakes to avoid include unclear statements, flowery marketing language, overuse of visuals, not using concrete data, overwhelming the audience with information, sending out editable files, and trying to sound more impressive than you are. By following these guidelines, founders can create a compelling pitch deck that effectively communicates their business's value proposition and increases the chances of securing funding or investment.