March 2025 Summaries

In an episode of Tractable, Kshitij Grover, co-founder and CTO of Orb, interviews Uma Chingunde, VP of Engineering at Render, about managing and optimizing complex technical platforms. The conversation focuses on the importance of continuous AI investment for reliability, the need for transparency and a blameless approach during critical incidents, and the impact of decision-making on operational changes. Render aims to be a modern cloud platform for developers, offering a platform as a service that abstracts away infrastructure complexities to enhance the developer experience. By bridging the gap between the resources provided by traditional cloud services and developers' desire for streamlined workflows, Render seeks to eliminate the undifferentiated toil engineers often face. Uma draws on her extensive experience, including her time at Stripe, to discuss strategies for running mission-critical workloads and serving diverse user segments effectively.

The "Engineering the Bottom Line" event series gathered experts from OpenAI, Lamini, Pinecone, and Orb to discuss strategies for conveying enterprise value through AI. The panelists highlighted crucial aspects of the AI stack, such as foundational models, training and inference, data, and pricing. They emphasized the importance of creativity and speed in AI development, noting that creative AI applications often yield better performance and adoption. Challenges such as pricing friction and the debate between open and closed AI models were discussed, with insights on optimizing AI products for profitability. The conversation also delved into topics like enterprise customization demands for AI models, the evolution of vector databases, and the need for flexible pricing strategies to adapt to the rapidly changing AI market.

In a podcast episode of Tractable, Kshitij Grover talks with Charity Majors, CTO of Honeycomb, about the evolution of observability tools and practices, transitioning from Observability 1.0 to 2.0. Observability 1.0 is characterized by multiple, disconnected sources of truth that rely heavily on engineers' intuition and experience for debugging, while Observability 2.0 provides a single source of truth with interconnected data allowing for more efficient problem-solving and shorter feedback loops. This shift enables engineers to derive metrics and traces from structured data, significantly speeding up processes such as identifying latency problems. Charity emphasizes that Observability 2.0 changes the way organizations develop code by making workflows more visible, reducing the time required to solve issues, and enhancing overall software engineering efficiency.

A panel discussion hosted by Orb and Baseten brought AI industry experts together to explore the multifaceted challenges of pricing, costs, and monetization in the AI sector. The conversation, moderated by Corinne Marie Riley from Greylock, included insights from leaders such as Matan-Paul Shetrit from Writer, Kshitij Grover from Orb, Joshua Ma from Airtable, Evan Conrad from San Francisco Compute, and Tuhin Srivastava from Baseten. The panel addressed the balance between innovation and revenue generation, emphasizing the importance of aligning AI capabilities with customer-perceived value and the necessity for AI startups to consider monetization strategies from the outset. As the AI market evolves, emerging trends in pricing models, like prepaid credits, value-based pricing, and hybrid models, were discussed, alongside the challenges of enterprise sales in the AI era and the build versus buy dilemma in AI development. The discussion concluded with a focus on the future of AI models, suggesting a strategic blend of generalized and specialized models to maximize customer value, and underscoring the need for AI companies to remain flexible and innovative in both technology and business approaches to harness the transformative potential of AI.

In an episode of the Tractable podcast, Kshitij Grover, co-founder and CTO of Orb, engages in a conversation with Erik, the founder and CEO of Modal, about Erik's technical journey and his experience in the tech industry. They explore Erik's early work experience at Spotify, where he was instrumental in developing a music recommendation system using data-driven techniques and overcoming the challenges of the nascent field of machine learning and AI. The discussion highlights the evolution of serverless technology and how Erik's company, Modal, simplifies cloud deployments, especially for data-intensive and AI workloads. Erik's past contributions, including open-source projects like Luigi and Annoy, reflect his innovative approach to solving complex data and machine learning problems, which has influenced his current work with Modal, a platform known for its scalability and ease of use in handling large workloads.

In an episode of the podcast Tractable, Orb CTO Kshitij Grover interviews Dane Knecht, the SVP of Emerging Technology and Incubation at Cloudflare, discussing Cloudflare's cutting-edge infrastructure, AI integration, and developer platform. The conversation highlights Cloudflare's expansive global network, which spans over 300 cities and 500 data centers, supporting 20% of the Internet, and its evolution from providing basic infrastructure protection to offering advanced performance solutions like WAF, DDoS, CDN, and caching. Furthermore, the discussion delves into Cloudflare's strategic shift to protect users and devices with products like 1.1.1.1 and warp, evolving into a comprehensive SaaS solution through zero-trust offerings. Knecht shares insights into Cloudflare's dual product engineering strategy that focuses on both revenue growth and pioneering innovative products that anticipate future market needs.

A panel discussion hosted by Orb and Simon-Kucher explored the intersection of AI, monetization, and future predictions, featuring insights from Naomi Pilosof Ionita, Joshua Bloom, and Sara Yamase, moderated by Michelle Verwest. The conversation highlighted the shift from the hype around AI in 2023 to its practical implementation in 2024, emphasizing the necessity for AI solutions to address real pain points and provide ROI while navigating the complexities of sales and legal reviews. Concerns around AI adoption include performance gains, data privacy, and integration, with a focus on transitioning from outdated seat-based pricing models to innovative monetization strategies such as activity-based and usage-based pricing. The panelists discussed the importance of balancing adoption and monetization, leveraging proprietary datasets for differentiation, and foresaw video as the next disruptive trend in AI. They also addressed the challenges of proving AI's value, maintaining margins, and the strategic significance of infrastructure investment, urging companies to be bold in their monetization approaches and capitalize on proprietary data to stand out in a competitive market.

In a recent episode of the Tractable podcast, Kshitij Grover, co-founder and CTO of Orb, converses with Can Duruk, co-founder and CTO of Felt, about the transformative potential of their collaborative mapping tool. Can shares his journey from working at various software companies, including Uber, to establishing Felt with his co-founder Sam in late 2020. Their platform emerged from a shared interest in mapping and visualization, motivated by their experiences with tools like Uber's internal mapping systems and Sam's previous work at Remix, a transit planning company. The discussion delves into the technical challenges of map rendering and the diverse applications of Felt's platform, as it redefines how geographic data is visualized and utilized across different industries, such as environmental investigation, sales charting, and more.

Orb collaborated with pricing consultancy Simon-Kucher to host a round-table discussion on effective monetization strategies, featuring insights from Orb Co-founder Alvaro Morales and Madhavan Ramanujam, author of "Monetizing Innovation." The discussion underscored the importance of product-market-pricing fit, flexibility, and segmentation in pricing models, emphasizing that businesses should avoid assuming a one-size-fits-all approach. Alvaro shared his experience with pricing at Asana, highlighting the challenges and revenue potential of strategic pricing adjustments, which led to the creation of Orb to alleviate related backend issues. Madhavan argued for a shift from seat-based pricing to more adaptive models, like usage-based or hybrid systems, especially in the AI-driven market where speed and strategic pricing are crucial for survival. The conversation also touched on the need for businesses to prioritize pricing strategies in tandem with product development to ensure sustainability and growth. Ultimately, the dialogue stressed the interconnectedness of acquisition, monetization, and retention, urging companies to balance these elements for long-term success.

Sentry, a leading error and performance monitoring tool used by over 90,000 companies including Atlassian, GitHub, and Disney+, has evolved from an open-source project initiated by David Cramer in 2008 to a critical component of the modern software stack. In an episode of the Tractable Podcast, David, the CTO of Sentry, discusses with Kshitij Grover, CTO at Orb, how Sentry grew from a simple error dashboard to a widely adopted tool through strategic decisions, technical challenges, and leadership styles. The conversation highlights Sentry's unique growth strategy influenced by its open-source origins, its adaptability to industry demands, and its approach to AI and brand campaigns, emphasizing how open-source models can drive market share growth. Despite its evolution, Sentry maintains its core functionality while expanding its product offerings to support a diverse range of frameworks and technologies, establishing itself as a well-recognized brand in the tech industry.

In an episode of the Tractable Podcast, Kshitij Grover interviews Neil Jagdish Patel, Co-founder and CEO of Axiom, a serverless logging tool designed for handling event data efficiently across various needs, including logging, observability, security, and product analytics. Axiom, used by companies like Prisma, M1 Finance, and Cal.com, is built on modern cloud technologies and specializes in storing event data with key-value pairs that include time, optimizing data ingest, storage, and query execution. Neil Patel discusses Axiom's potential as a central tool for organizational events, its emphasis on customer experience, and its product philosophy, which supports various data ingest methods. He also highlights its unique approach to query planning and execution, as well as its AWS S3-based storage architecture. The conversation delves into Axiom's flexibility and trade-offs, its commitment to accessibility, and future product launches, positioning it as a versatile solution for event management and data processing.

In a podcast episode of Tractable, Kshitij Grover and Richie Artoul discuss the development and advantages of WarpStream, a Kafka-compatible streaming platform designed to be more cost-effective and manageable than traditional Kafka systems. Built on Richie's experience with large-scale data systems at Datadog and Uber, WarpStream offers a stateless architecture that reduces inter-AZ costs and simplifies management while maintaining generally acceptable latency levels. The conversation also highlights WarpStream's integration with S3 and its serverless pricing model, which positions it as a critical infrastructure component for other companies. Richie shares his background in distributed storage and observability technology, recounting his work on Uber's internal metrics platform, M3, and explaining how the need for a cost-effective solution for massive data throughput inspired the development of WarpStream.

In a podcast episode of Tractable, Kshitij Grover, CTO and Co-founder of Orb, discusses with Gwen Shapira, Co-founder of Nile, the complexities and innovations in multi-tenancy and serverless Postgres offerings for SaaS applications. They focus on Nile's approach to virtual tenant databases and the challenge of balancing flexibility, isolation, and resource allocation when managing mixed data blocks for diverse customer needs. Gwen shares insights from her extensive career in software engineering and data architecture, including her experience with Oracle, MySQL, Hadoop, Kafka, and Confluent. The conversation highlights the central role of customer data models in B2B SaaS and the ongoing trade-offs between dedicating databases to individual customers versus employing a multi-tenancy approach, emphasizing the evolving demands on reliability and scalability in modern database solutions.