Tigris Data is a developer-first infrastructure company that builds a globally distributed, S3‑compatible object storage service optimized for low‑latency access and AI/ML workloads, founded in 2022 by former Uber storage engineers who previously operated Uber’s global storage platform[2][3]. Tigris positions itself as the storage foundation for AI-era applications by offering global data placement, low-latency access for small objects (e.g., embeddings), multi-cloud mobility, and tooling to migrate workloads from other providers[5][4].
High‑Level Overview
- Mission: Tigris aims to provide a storage foundation that “just works” for developers — global, multi‑cloud, and optimized for AI and real‑time applications so teams can focus on product rather than storage plumbing[2][5].
- Investment philosophy / Key sectors / Impact on the startup ecosystem: (Not applicable — Tigris is a portfolio company / product company; funding rounds and investors include a seed led by a16z and later investments referenced in company announcements)[7][5].
- What product it builds: Tigris offers a globally distributed, S3‑compatible object storage service with features like access‑based rebalancing, caching, and tiered storage optimized for AI and small‑object workloads[8][5].
- Who it serves: Developers, AI startups, next‑gen cloud platforms, and any teams needing low‑latency, globally distributed storage for training data, embeddings, model serving, logs, and media[5][4].
- What problem it solves: It reduces storage latency and operational complexity by automatically placing and caching data near users, avoiding heavy lift migrations, and enabling efficient handling of small files common in AI workloads[8][4].
- Growth momentum: Tigris has raised venture funding (seed led by a16z and later rounds), reports thousands of customers and plans to expand hardware regions internationally, and secured a $25M raise in 2025 to scale its AI‑optimized storage service and data center footprint[7][5][4].
Origin Story
- Founding year and founders: Tigris was founded in 2022 by Ovais Tariq (CEO), Himank Chaudhary (CTO), and Yevgeniy Firsov (Chief Architect), who previously worked together for nearly six years on Uber’s global storage platform, operating high‑scale systems such as Docstore, Herb, and DBEvents[2][3].
- How the idea emerged: The founders’ Uber experience highlighted the value of architecture and operational discipline at scale; they started Tigris to bring the same reliability and developer freedom to other teams, especially as AI creates new storage demands[2][5].
- Early traction / pivotal moments: Early investor support included Andreessen Horowitz for the seed round and later endorsements from firms like Spark Capital; by 2025 the company announced a $25M raise and publicized thousands of customers and multi‑region hosting to support AI workloads[7][5][4].
Core Differentiators
- Product differentiators:
- Globally distributed, S3‑compatible object storage with *access‑based rebalancing* that automatically caches and relocates hot data to regions where it’s accessed[8].
- Optimizations for small objects and embeddings to reduce retrieval latency compared with typical S3-like systems[4].
- Tiered storage options (standard, infrequent, and archive tiers) to balance cost and performance[4].
- Developer experience:
- S3 API compatibility lets existing tools and libraries work with minimal changes[8].
- Migration tooling such as “shadow buckets” to gradually copy hot data and avoid big‑bang migrations[4].
- Performance and operational model:
- Uses LSM‑style data structures and caching strategies to improve small‑object access patterns and overall latency[4].
- Focus on running their own hardware and data centers to control latency and avoid some cloud egress penalties, with plans for more global regions[2][5][4].
- Community/ecosystem:
- Positions itself as open and developer‑friendly and is being adopted by next‑gen developer cloud platforms and AI startups listed as customers and partners[6][5].
Role in the Broader Tech Landscape
- Trend they are riding: The surge in AI/ML and embedding‑driven applications is driving need for storage that handles massive numbers of small objects with low latency and global access patterns, which Tigris explicitly targets[4][5].
- Why timing matters: As AI moves from experimental to production, latency and data locality matter more for user experience and cost; multi‑cloud/multi‑region data strategies and egress costs are growing concerns that create demand for alternatives to single‑cloud storage[5].
- Market forces in their favor: Rising volumes of training and inference data, the proliferation of generative AI apps, and interest in best‑of‑breed infrastructure providers rather than vertically bundled cloud stacks support adoption of specialized storage layers[5].
- Influence on the ecosystem: By enabling lower‑latency global access and easier migration, Tigris helps AI startups and developer‑cloud providers focus on product features rather than storage operations, and it pressures incumbent cloud providers to improve small‑object and global‑distribution performance[5][4].
Quick Take & Future Outlook
- What’s next: Tigris is likely to continue expanding its region footprint and hardware presence (plans announced for London, Frankfurt, Singapore), deepen AI optimizations (embeddings, caching, storage tiers), and broaden integrations with ML platforms and developer clouds[4][5].
- Trends that will shape their journey: Continued growth of embedding usage in retrieval‑augmented generation (RAG), wider deployment of latency‑sensitive AI services, and corporate focus on multi‑cloud portability and egress costs will drive demand for specialized storage[4][5].
- Potential evolution: If Tigris successfully scales and keeps pricing/performance competitive, it could become a common storage layer for AI stacks and developer clouds; conversely, it must sustain operational excellence and ecosystem partnerships to fend off competition from major cloud providers and other niche storage startups[5][4].
Quick take: Tigris is a technically credible, founder‑led storage startup with direct experience operating at extreme scale; its focus on low‑latency, globally distributed storage for AI use cases addresses a clear market need, and its recent funding and customer traction suggest meaningful momentum as AI workloads continue to proliferate[2][5][4].
