The AI customer service agent built to resolve the messy tickets chatbots hand back to humans - refunds, partial-order exchanges, address changes - end to end.
Tymely's green wordmark, its lowercase "m" resting on three dots that echo a live-chat "typing" bubble. New York & Israel, founded 2021.
Every online shopper knows the feeling. The order arrived in the wrong size, the return window is confusing, and the "help" chat loops back to a menu that never had the right option. For a decade, retailers threw chatbots at the problem and mostly automated the easy questions - "where is my order?" - while the tickets that actually mattered still landed in a human queue.
Tymely, a company founded in 2021 and split between New York and Israel, built its product around the opposite end of that queue. Its AI agent is designed to resolve the complicated cases end to end: exchanging part of an order, changing a shipping address mid-transit, issuing a refund with the right exception applied. These are the tickets that require reading order data, taking real actions across a retailer's systems, and using judgment - the work that generic chatbots quietly escalate.
The company frames its own origin in blunt terms. "Customer service today is broken," co-founder and CEO Ohad Rozen has said. "Many businesses are losing a lot of money because of poor customer service." Tymely's answer is not a smarter FAQ. It is an attempt to clone the reasoning of a brand's best agents and put that reasoning to work across email and chat, around the clock.
Figures are company-reported and drawn from public case studies and press. Treat as approximate.
By Tymely's own telling, the value in support hides in the hard cases - the roughly 80% of inquiries that carry an exception, a judgment call, or a multi-step action. A case study the company promotes describes a retailer whose customer-satisfaction score rose from 3.9 to 4.6 while response times dropped sharply after deployment. The claim worth noting is directional: speed and quality stopped behaving like a tradeoff once the AI could reason through a ticket rather than pattern-match a canned reply.
We created a solution that harnesses the power of human verification and human conscience and the brilliance of AI technology.
Most tools marketed as "AI for customer service" hand the ticket back to a person the moment it gets complicated. Tymely positions itself against that reflex. Its stated design rule is simple: if the AI cannot resolve a ticket correctly, it should not send a reply at all - which is the company's explanation for why its case studies talk about satisfaction rising rather than falling.
An AI customer service agent that resolves complex eCommerce and retail tickets across email and chat - refunds, partial-order exchanges, shipping-address changes and other multi-step actions - behind operational guardrails and real-time verification.
Reverse-engineers how a brand's top agents reason through tickets, mapping their actual decision logic and undocumented knowledge so the AI can replicate human-grade judgment on edge cases.
Plugs into major CRM and helpdesk platforms so the agent can read order data and take actions inside a retailer's existing systems rather than living in a silo.
Positioned for enterprise deployment with compliance and data-security controls, so an AI issuing refunds and changing orders operates inside a brand's risk boundaries.
Tymely sells to mid-market and enterprise eCommerce, direct-to-consumer and retail brands - companies whose support headcount grows with every order spike and seasonal peak. Publicly cited customers include a mix of national and D2C names.
Tymely operates as B2B SaaS - effectively an outsourced AI contact center. Rather than positioning as a self-serve chatbot add-on, it pitches itself as a replacement for the offshore and BPO support labor that brands lean on today. That framing changes the buyer and the math: the comparison is not "chatbot vs. no chatbot" but "AI agent vs. a team of humans," where covering the complex majority of tickets is what justifies the switch.
In the broader market, Tymely sits among a crop of AI-native support agents - names like Sierra, Decagon, Ada, Forethought and Gorgias's AI agent - all racing toward autonomous resolution. Tymely's differentiator, at least in its own positioning, is the emphasis on the hardest, action-heavy tickets and on not sending a reply it cannot stand behind.
An NLP and deep-learning researcher who worked as a research scientist at Amazon and Yahoo and previously led Toonimo. The product is, in effect, his research applied to support queues.
A veteran engineering leader - previously VP of Engineering at Amenity Analytics - with roughly two decades building and shipping technology products.
The company's edge is less a single feature than the founders' pedigree: applied natural-language research aimed squarely at the unglamorous, high-volume reality of retail support. Their thesis is that the most valuable data in a support organization - which exceptions to make, when to bend a policy, how to read an angry email - was never written down, and that extracting it is the real product.
| Round | Amount | Date | Key Investors |
|---|---|---|---|
| Series A | $7.0M | Sep 2022 | Hetz Ventures, DESCOvery (D.E. Shaw Group) |
| Later round (reported) | ~$8.25M | 2024 | Alea Capital Partners, Arrow Venture Partners, 97212 Ventures, ICON, DESCOvery |
Total reported funding ~$15.25M. Round labels and dates vary across public data providers - treat later-round details as approximate.
Ohad Rozen and Roy Penn start the company to fix broken eCommerce customer service with NLP.
Round led by Hetz Ventures and DESCOvery to scale the NLU technology and build a sales team.
Publishes case studies and adds national retail brands including Victoria's Secret and Paul Fredrick.
Reported to raise further capital, bringing total funding to roughly $15.25M.
It provides an AI customer service agent that resolves complex eCommerce and retail support tickets end-to-end across email and chat - including refunds, partial-order exchanges and address changes.
Ohad Rozen (CEO) and Roy Penn (CTO) founded it in 2021. Both have backgrounds in NLP and AI research, including roles at Amazon and Yahoo.
Instead of deflecting FAQs, Tymely reverse-engineers how a brand's best agents reason and uses guardrails and verification to fully resolve complex, action-heavy tickets without sending replies it can't stand behind.
eCommerce, D2C and retail brands. Publicly cited names include Victoria's Secret, Paul Fredrick, Dailylook, Kuru, Nanit and Underoutfit.
A $7M Series A in 2022 led by Hetz Ventures and DESCOvery, with total reported funding of roughly $15.25M.
No public YouTube channel, product-demo video, Twitter/X or Instagram account was verified for Tymely at the time of writing. Contact: info@tymely.ai