Avent: Product, team and how it was built
Avent (AI agents that prepare quotes and enter orders for industrial distributors and manufacturers, United States): Live for industrial distributors, November 2026. Evidence grade E: no revenue figure was checked. Team: 1 founder, Abhay Kalra; 5 people on the Y Combinator page in November 2026. Launched 15 August 2025; still running.Added 7 November 2026
- Outcome
- Live for industrial distributors, November 2026
- Evidence
- E no revenue figure was checked
- Launch
- 15 August 2025
- Status
- Still running
- Team
- 1 founder, Abhay Kalra; 5 people on the Y Combinator page in November 2026
- Outside funding
- Y Combinator (Summer 2025 batch); amount not published
- Main channel
- Not stated
What is not verified
No revenue figure was found. This page looked for a TrustMRR profile (there is none in its list), a founder interview with figures and a filing, and found none. The trade magazine article by Modern Distribution Management could not be opened (403) and is known only from the company's summary of it. The figure of more than US$200,000 worth of quotes a day is the founder's statement of August 2025, with no second source and not repeated on the pages read in November 2026; it is the value of customers' quotes, not Avent's revenue. A reader could look for a press interview with figures or a customer case study. This story is grade E: the business is real and its product, price and customers are described from public pages, but no revenue figure was checked. Aiolum shows it as an idea worth studying, not as proof that it makes money.
Avent sells AI agents, programs that use a large language model to read a request and then take steps in other software, to industrial distributors, the wholesale companies that buy parts from manufacturers and resell them to other businesses. When a buyer emails a request for quote (RFQ), Avent matches the items to the distributor's catalog, applies its pricing rules and prepares the quote for approval. When the purchase order (PO) follows, it checks it against the quote and creates the sales order in the distributor's ERP (enterprise resource planning system, the database that holds stock, prices and orders).
Abhay Kalra founded it in San Francisco in 2025 and took it through Y Combinator's Summer 2025 batch. He studied electrical engineering and computer science at UC Berkeley and grew up around his family's electrical and industrial distribution business in India. In the launch post of 15 August 2025 he wrote that Avent generated more than US$200,000 worth of quotes every day for distributors in North America and the UK. In November 2026 Y Combinator's page lists 5 people.
No revenue figure was found, so this is a grade E story and its headline is the product's status: live, sold after a demo, with no public prices. The US$200,000 a day that the founder stated is the value of quotes that customers sent to their own buyers, not Avent's revenue.

How much revenue does Avent make?
Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.
| Fact | Value | Grade | How we know | Sources |
|---|---|---|---|---|
| Status, 7 November 2026 | live; sold after a demo, with paid terms and no public prices | C | The home page offers "Get a Demo" and "Contact Sales" and links to a login at app.aventindustrial.com. The terms of service have a section "Payment Terms" for "our paid Services". The /pricing page returns 404. | [3, 12] |
| Year founded | 2025, in San Francisco | B | Y Combinator's company page, which lists the batch as Summer 2025. | [2] |
| Founder | Abhay Kalra, founder and CEO | C | Y Combinator's page lists him as the only founder; its bio says he studied Electrical Engineering and Computer Science at UC Berkeley and was an operator in manufacturing and industrial distribution. The About page is a letter signed by him. | [2, 4] |
| Team size, November 2026 | 5 | C | Team size field on the Y Combinator page, filled in by the company. The careers page says "We're a small team working in person in San Francisco" and lists 2 open engineering roles. | [2, 10] |
| Public launch | 15 August 2025 | C | Y Combinator launch post "Avent – AI Agents for Industrial Commerce", dated 15 August 2025, with 117 votes when read on 7 November 2026. The company's blog post of the same day announces the launch out of Y Combinator. | [1, 5] |
| Value of quotes generated, August 2025 | more than US$200,000 a day, as the founder stated | C | Launch post: "Avent generates over $200,000 worth of quotes every day for distributors across North America and the UK." This is the value of customers' quotes, not Avent's revenue, and not all quotes become orders. Not verified. | [1] |
| Prices, November 2026 | not published | C | No pricing page; sales by demo and contact form. | [3] |
| AI model providers, October 2026 | Anthropic for AI features; OpenAI for search embeddings | C | Privacy policy, last updated 26 October 2026: "Relevant content may be processed by Anthropic for AI features" and Knowledge Base content "may be processed by OpenAI to create search embeddings". Custom code in workflows runs in an E2B cloud sandbox. | [11] |
| ERP systems named | SAP, Oracle, Epicor; the home page also shows Infor, NetSuite, Dynamics 365 and Salesforce | C | Launch blog post: "SAP, Oracle, Epicor, and more". The logos are in the home page's integrations section, not presented as customers. | [5, 3] |
| Outside funding | Y Combinator, Summer 2025 batch; amount not published | B | Batch shown on Y Combinator's page. No round or amount is published on the pages read. | [2] |
| Time from launch to the November 2026 status | 15 months | calc | Calculated: August 2025 to November 2026. | [1, 3] |
How to read this table: a grade C value is the company's or the founder's own statement on its site or in its Y Combinator launch post. A grade B value comes from Y Combinator's own company page. A "calc" value (grade X) was computed on this page. No value is grade A, and no revenue figure was found.
| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Platform: quoting, order entry, inbox management, knowledge base, voice | not published | demo, then a contract with sales |
Every page read ends with "Contact Sales" or "Get a Demo", which opens a discovery call booked through the founder's calendar link. The terms of service, last updated 26 October 2026, say that customers who subscribe to the paid Services pay the fees "presented at the time of purchase" and that payments are non-refundable unless stated otherwise. No plan, price or seat count is published.
When did Avent launch, and how did it grow?
From a Y Combinator batch to a quote-to-cash platform shown at a distributors' trade show, in 15 months. Each entry carries its grade.
- 2025BAvent is founded in San Francisco and joins Y Combinator's Summer 2025 batch, with Abhay Kalra as its only founder. [2]
- 15 August 2025CLaunch post on Y Combinator: an AI inside sales rep that quotes, enters orders and searches products for distributors, and more than US$200,000 worth of quotes a day, the founder wrote. He asks for introductions to distributors, manufacturers and independent manufacturers' reps. [1]
- 15 August 2025CThe company's blog announces the launch and a feature on the founder in the trade magazine Modern Distribution Management, which describes quotes, order entry and product search. [5, 6]
- April 2026CAvent announces a booth at the ISA26 trade show on 21 and 22 April, with two workshops, live demos, a workflow builder, vendor communication tools and a new pricing engine. [7]
- 26 October 2026CThe privacy policy and terms are updated. The policy names Anthropic, OpenAI and E2B, and sets rules for Gmail and Google Sheets access and for voice data. [11, 12]
- 7 November 2026CThe home page sells an "AI Workforce for Distribution" with the Forge workflow editor and the Arc copilot. Y Combinator lists 5 people, and 2 founding engineer roles are open. [3, 2, 10]
Why did Avent work?
Four causes, in the order they mattered. Each one is a reusable lesson.
A problem the founder saw from inside
The About page tells how the founder's father built a distribution business and how every plan for growth ended at hiring, because the sales team spent its day on quotes and order entry. The product does exactly those two tasks. In the launch post he names his customers in the trade's own terms: distributors, manufacturers and independent manufacturers' reps.
The model reads, the rules decide
In the home page example, Avent's AI matches a request for 12 pneumatic cylinders to a catalog item, and a pricing rule from the ERP price list sets the price; a person approves the quote. The customer chooses at which steps AI acts, and the example permissions let the quoting agent prepare and run work but not approve it. When a person corrects information, the dependent steps continue from the corrected version, the home page says.
On top of the ERP, not instead of it
Distributors keep their SAP, Oracle or Epicor system. Avent reads email, writes sales orders into the ERP and lets a customer "Start with one process, then expand on the same platform—without replacing your ERP", in the home page's words. That lowers the cost of saying yes for a company that runs on decades-old systems.
Sold by demo inside the trade
There is no self-service sign-up. Every page leads to a discovery call, and the company went to the ISA26 trade show with a booth and two workshops. No source says how most customers found it.
How it was built
These details come from the company's own pages and privacy policy (grade C) and from the page source of its website (grade B). [11, 5, 3, 10]
- Language models: the privacy policy says content may be processed by Anthropic for AI features and by OpenAI to create search embeddings for the Knowledge Base.
- Custom code that a customer adds to a workflow runs in an E2B cloud sandbox, the privacy policy says.
- Connects to Gmail and Google Sheets: Avent can read, search, send and organize email and read and update chosen sheets for the configured workflows.
- Integrations with ERP and CRM systems: SAP, Oracle, Epicor in the launch post; Infor, NetSuite, Microsoft Dynamics 365 and Salesforce logos on the home page.
- The website is a Next.js application, visible in its page source; the product runs at app.aventindustrial.com.
- Demos are booked through a Cal.com link in the founder's name, and jobs are posted on Ashby. The home page says the company is SOC 2 compliant and shows an AICPA SOC 2 seal.
Tools named in the sources
Each one opens its page in the Tools tab, with every story that uses it.
Who tried this before Avent, and why did they fail?
Three companies in our database of failed startups promised to take the manual work out of business buying, ordering or quoting and failed. The values below are from that database and are not independently verified.
Kinnek
The promiseA marketplace that turns a small business's search for suppliers and quotes into one streamlined process.
Why it diedAmazon Business and Alibaba already had the supplier networks, and the platform could not fit the different needs of many small buyers.
TradeGecko
The promiseInventory and order management for small wholesale and online sellers, with integrations to their other systems.
Why it diedAfter Intuit bought it, the focus moved to QuickBooks, and keeping many integrations up to date cost more than its resources allowed.
Shipitwise
The promiseAutomated freight quotes and logistics for small companies without a shipping department.
Why it diedIt depended on third-party logistics partners and spent its money on complex integrations, and too few customers adopted it.
The failed startups come from our database of 1,749 dead startups, retold here in our own words. Its entries are AI-assisted summaries of public sources, so these cards carry no evidence grade.
What changed
Kinnek asked buyers to move to a new marketplace, and TradeGecko and Shipitwise asked businesses to run orders or freight through a new system that had to be connected to everything else. Each one spent heavily on integrations or on building a network before customers came. Avent does not ask the buyer to change anything: the request still arrives by email, and the order still ends up in the distributor's own ERP.
What earlier software could not do well was read the request itself: a free-text email, a PDF or a part number in a customer's own shorthand. Avent's privacy policy says it uses Anthropic's models for its AI features, so the company rents that reading ability and spends its own work on pricing rules, approvals and ERP connections. Integrations are still the expensive part, as they were for TradeGecko and Shipitwise.
Can one person copy Avent now?
The method, yes. The trade knowledge, only from inside the trade.
A small team can now build an agent that reads free-text requests and writes orders into an old system, because the reading is done by rented models. What is hard to copy is knowing how distributors price, which exceptions matter and whom to call; the founder brought that from his family's business, by his own account.
Where it is still open: Other trades that sell by request for quote: electrical and plumbing wholesalers, building-materials suppliers, chemical and plastics distributors, and manufacturers' reps outside North America and the UK.
- 1Pick a back-office task in a trade where requests arrive as free text, such as quotes, purchase orders or freight requests.
- 2Let the model read and match, but keep prices and approvals in rules that the customer can see and edit.
- 3Write into the customer's existing ERP instead of asking them to change systems.
- 4Sell through the trade's own channels: introductions, trade shows and discovery calls.
What could go wrong for Avent?
The risks are the mirror image of the causes above.
What is not verified
No revenue figure was found. This page looked for a TrustMRR profile (there is none in its list), a founder interview with figures and a filing, and found none. The trade magazine article by Modern Distribution Management could not be opened (403) and is known only from the company's summary of it. The figure of more than US$200,000 worth of quotes a day is the founder's statement of August 2025, with no second source and not repeated on the pages read in November 2026; it is the value of customers' quotes, not Avent's revenue. A reader could look for a press interview with figures or a customer case study.
Calls, email and data use
The voice feature listens to sales calls and drafts quotes, and the privacy policy says "You are responsible for ensuring you have appropriate consent from all parties before submitting or recording conversations". The Gmail connection lets Avent read and send email for configured workflows. The home page says "No shared training between customers"; the privacy policy lists "Generate and train machine learning models to improve our AI systems" among its uses and says aggregated, anonymized voice data may improve its models, with Google user data excluded.
Competition and model providers
On this site the closest stories are Lark, AI agents for wholesale distributors, Anchr, order intake at food distributors, and LineSight, order entry for metal service centers; Avent aims at industrial parts distributors and their manufacturers. The privacy policy names Anthropic for AI features and OpenAI for search embeddings, so changes in their prices or terms would reach Avent's costs.
What the pages do not show
The pages read name no customer and show no testimonials; the quote example on the home page is labelled "Sample data". The home page says the company is SOC 2 compliant and shows an AICPA SOC 2 seal, but does not name the auditor, the report type or a trust center. Prices are not published.
Sources
The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.
- C Y Combinator launch post, 15 August 2025: Avent – AI Agents for Industrial CommerceRead in full, from the post data in the page. Founder's own post. Source for the product, the quote-value claim, the founder's story and the vote count.
- B Y Combinator company page: AventRead the company data in the page source: founded 2025, batch Summer 2025, team size 5, San Francisco, founder Abhay Kalra and his bio.
- C Avent home page: AI Workforce for Distribution, read 7 November 2026Read in full. Source for the product, the Forge and Arc features, the sample quote, the integrations, the SOC 2 claim and the demo links.
- C Avent About page: letter from Abhay Kalra, read 7 November 2026Read in full. Source for the family business and the hiring problem.
- C Avent blog, 15 August 2025: Avent Launches Out of Y Combinator to Reimagine Sales for Distributors and SuppliersRead in full. Source for the launch, the ERP systems named and the hiring plans.
- C Avent blog, 15 August 2025: MDM Speaks with Our Founder: How a Second-Generation Distributor Built AventRead in full. The company's summary of a Modern Distribution Management article, which could not be opened (403). Source for the founder's background in India and at UC Berkeley.
- C Avent blog, 14 April 2026: Avent Is Heading to ISA26 — Come See Us at Booth #1009Read in full. Source for the trade show, the workshops and the features shown.
- C Avent solution page: Voice Agents, read 7 November 2026Read in full; the FAQ answers load by script and were not read. Source for the call notes and post-call quotes.
- C Avent solution page: Quote Automation, read 7 November 2026Read in full; the FAQ answers load by script and were not read. Source for shorthand matching, customer pricing and sourcing.
- C Avent careers page, read 7 November 2026Read in full. Source for the small in-person team and the 2 open roles on Ashby.
- C Avent privacy policy, last updated 26 October 2026Read in full. Source for the model providers, the E2B sandbox, the Gmail and Google Sheets access, and the voice data and consent sections.
- C Avent terms of service, last updated 26 October 2026Read the sections on payment terms and the legal entity, Avent Industrial Inc., under Delaware law.
- C Database of failed startups, local exportLocal export. AI-written summaries, not independently verified. Source for the failed predecessors.
Run this company? Send corrections or a verified figure to aiolum@livyden.com