Jigsawstack is a technology company.
Jigsawstack has raised $1.0M across 1 funding round.
Jigsawstack has raised $1.0M in total across 1 funding round.
Jigsawstack has raised $1.0M in total across 1 funding round.
Jigsawstack's investors include Ada Ventures, Andreessen Horowitz, Atomico, Creator Fund, Double Prime LLP, Russell Smith.
JigsawStack is a Singapore-based technology startup that builds a developer-friendly AI SDK featuring purpose-built small AI models for backend tasks like web scraping, vOCR (vision OCR), object detection, translation, speech-to-text, and more.[1][2][3] It serves backend developers and engineers at scale, solving chronic pain points such as frequent maintenance of unreliable services (e.g., scraping breaking due to site changes), high costs, poor performance, and lack of consistency from general-purpose large models.[2] The platform enables seamless integration via fully typed SDKs in languages like JavaScript, Python, and others, with serverless scaling to billions of inferences in under 200ms, pay-per-use pricing, global GPU support, and real-time analytics—trusted by builders for fast, reliable results in mission-critical applications from healthcare to military.[1][2]
JigsawStack recently closed a US$1M pre-seed round led by UK-based Ada Ventures, signaling strong early momentum in the AI infrastructure space.[4] Its growth is driven by a community of AI engineers shipping faster, with models trained on custom global data for 160+ languages and specialized tasks, outperforming bloated alternatives from cloud giants.[1][2]
JigsawStack was founded by Yoeven Khemlani, who identified a massive inefficiency in backend development: engineers spend ~80% of their time maintaining fragile services like scraping, OCR, and translation, which break frequently due to layout changes, IP issues, or costs.[2][3] In the AI era, Khemlani questioned why no simple SDK existed to automate these with adaptive, high-performance models—especially as large cloud providers offered outdated, expensive, hallucination-prone solutions with terrible developer experience.[2]
The idea emerged from this frustration, leading to the creation of small, specialized models combining traditional ML bases, compact language models, and vast custom-tagged data for tasks like vOCR and forecasting.[2][3] Pivotal early traction came via a podcast appearance in July 2025, highlighting its suite for collaborative knowledge work, followed by the US$1M pre-seed in March 2025 to scale AI deployment.[3][4] This bootstrapped evolution humanizes a team focused on DX-first principles for reliable backend AI.[2]
JigsawStack stands out in the crowded AI API market through these key strengths:
JigsawStack rides the small models revolution, capitalizing on the shift from resource-hungry LLMs to efficient, task-specific AI that fits backend constraints—perfect timing as enterprises demand reliable, cost-effective automation amid AI hype fatigue.[2][3] Market forces like exploding backend service needs (scraping, OCR in pipelines), GPU shortages, and scaling pains favor its lightweight approach, enabling billions of inferences without infrastructure headaches.[1][2]
It influences the ecosystem by democratizing high-performance AI for developers, reducing maintenance burdens and fostering faster innovation in research, strategy, and ops-heavy sectors; as a Singapore startup with global reach, it challenges US/Cloud giants, accelerating adoption in emerging markets.[3][4]
JigsawStack is poised for rapid expansion post-pre-seed, likely targeting Series A in 2026 by growing its model catalog (e.g., more multimodal/niche tools) and enterprise wins in regulated verticals.[4] Trends like edge AI, agentic workflows, and multimodal data will amplify its edge, as devs prioritize consistency over generality amid rising API costs. Its influence could evolve into a backend AI standard, much like Stripe for payments—transforming "high-maintenance" tasks into plug-and-play primitives, ultimately redefining scalable AI deployment from the ground up.[1][2]
Jigsawstack has raised $1.0M across 1 funding round. Most recently, it raised $1.0M Seed in February 2025.
| Date | Round | Lead Investors | Other Investors |
|---|---|---|---|
| Feb 1, 2025 | $1.0M Seed | Ada Ventures, Andreessen Horowitz, Atomico, Creator Fund, Double Prime LLP, Russell Smith |