Aiolum Browse stories
All stories
AI assistant for sales representatives · United States

Aida: Revenue, sale and how it was built

Small teamFundedExitB2BSales

Aida (AI assistant for sales representatives, United States): Acquired by Harvey, October 2026. Evidence grade E: no revenue figure was checked. Team: 2 co-founders at the start in 2024 (Piyush Prahladka and Atish Das Sarma); Dealroom lists 39 employees (read 8 November 2026); the team joined Harvey in October 2026. Launched 2024; sold 33 months after launch.Added 8 November 2026

Outcome
Acquired by Harvey, October 2026
Evidence
E no revenue figure was checked
Launch
2024
Status
Sold, 33 months after launch
Team
2 co-founders at the start in 2024 (Piyush Prahladka and Atish Das Sarma); Dealroom lists 39 employees (read 8 November 2026); the team joined Harvey in October 2026
Outside funding
Backed by 8VC, CRV and Firebolt and operator angels, the company said; Dealroom records a US$11M seed round in October 2025
Main channel
Not stated
No revenue figure

What is not verified

No revenue figure was found, and neither Harvey nor Aida disclosed the price. No payment-processor profile, founder interview with figures or filing was found. The usage figures (above 80% weekly active usage, 67% faster sales cycles, 2.5 times quota capacity, three times the output of one rep in beta) are the company's own claims. The headcount of 39 comes only from Dealroom's directory, read on 8 November 2026, after the sale, and the founding month in 2024 is not published. Dealroom's funding record (a US$11M seed round in October 2025 from Ascolta Ventures) does not name the investors that Aida names (8VC, CRV and Firebolt), so it may be incomplete. Harvey does not say whether Aida stays on sale to sales teams. A reader could look for a founder interview or Harvey's later product notes. 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.

Aida is an assistant for sales representatives that the company called an AI-native sales execution platform: software built around a language model from the start, which does the follow-up work of a sale and does not only record it. It connects a rep's email, calendar, call recordings and CRM, the customer relationship management system in which a sales team records its accounts and deals, and then drafts follow-up emails, updates deal records and warns when a buyer goes silent. In July 2026 it added a Digital Twin, the company's name for a model of one rep's own writing style and habits, learned from that rep's sent emails and calls.

Piyush Prahladka, who co-founded the enterprise search company Glean, and Atish Das Sarma, who had led engineering, data science and growth teams at Patreon, Apple and X, founded Aida in Palo Alto in 2024, according to the company's press release and Dealroom. Customers sign an Order Form and pay subscription fees under the terms of service; Eve Legal, a legal AI company, became a customer after a seven-month design partnership. On 29 October 2026 Harvey announced that it had acquired Aida and that the founders and their team of search and machine learning engineers would join its product and engineering organization.

Neither company disclosed a price, and no source gives a revenue figure, so this page is grade E: the buyer confirms the sale, but no business number is verified. Dealroom lists 39 employees, so this is a two-founder start that grew after venture funding, not a one-person business.

Aida: Acquired by Harvey, October 2026. AI assistant for sales representatives, United States, started small, grew after funding, evidence grade E.
The story in one card. Share it with the address of this page.
01

How much did Aida make, and what was it sold for?

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.

A two-founder sales AI company from 2024 that grew to about 39 people with venture funding and was bought by Harvey in October 2026; price and revenue not disclosed
FactValueGradeHow we knowSources
OutcomeAcquired by Harvey, announced 29 October 2026AHarvey's announcement: "Harvey today announced its acquisition of Aida". The co-founders and their team of search and machine learning engineers join Harvey's product and engineering organization.[1]
Sale pricenot disclosedBDealroom News: Harvey acquired Aida "for an undisclosed sum". Harvey's announcement gives no price.[2, 1]
Place in Harvey's acquisitionsthe fifth of 2026AAfter Benchmark in July and a deal in September that the announcement writes as "Guardails"; this site's Guardrails AI story covers the September deal.[1, 2]
Founded2024, in Palo Alto, CaliforniaCAida's press release of 15 July 2026 ("Founded in 2024 and headquartered in Palo Alto") and Dealroom's profile ("Founded 2024"). The month is not published.[4, 3]
FoundersPiyush Prahladka (co-founder and CEO) and Atish Das Sarma (co-founder and CTO)ABoth named by the buyer. Prahladka wrote on the company page that he worked on search at Google, ads on Maps and a transit business at Uber before he co-founded Glean. Das Sarma led engineering, data science and growth teams at Patreon, Apple and X, Harvey wrote.[1, 5]
Team39 employees in Dealroom's profileDRead 8 November 2026, after the sale. Dealroom also shows team growth of 105.3% in one year. The company page says "Small team, on purpose" and lists offices in Palo Alto and Gurugram, India. No other source gives a headcount.[3, 5]
Funding8VC, CRV and Firebolt, with operator angels; a US$11M seed round in October 2025 per DealroomCThe investors are named in Aida's press release and on its company page. The amount and date come only from Dealroom's profile, which lists Ascolta Ventures as the investor in that round (grade D).[4, 5, 3]
Paid productsubscription fees under an Order FormCTerms of service, section 10: "Customer will pay Company subscription fees for the Services in accordance with the commercial terms set forth in the applicable Order Form".[7]
Named customersEve Legal, WorkWave, Synctera, Straiker and TeamOhanaCPress release of 15 July 2026, which also says that Aida "has converted dozens of enterprise sales teams off legacy platforms".[4]
Eve Legala seven-month design partnership, then a paying customerCAida's case study. Eve Legal's founding account executive, Mercedes Quinonez, manages more than 100 customers and more than 100 prospects in Aida, the case study says.[6]
Usageabove 80%CThe company's claim in its July 2026 press release, set against the "sub-20 percent license utilization it says is typical of incumbent tools". Nobody outside the company has checked it.[4]
Time from founding to the saleat most 33 monthscalcCalculated: January 2024, the earliest month the founding year allows, to October 2026. It is 22 months if Aida was founded in December 2024.[4, 1]

How to read this table: grade A is the buyer's own announcement, grade B is an independent news report, grade C is the company's own press release or page, grade D is a directory record with no second source. A "calc" value was computed on this page. No row gives a price or a revenue amount, because no source states one.

Aida publishes no prices (read 8 November 2026)
PlanPer month, billed yearly (US$)Per month, billed monthly (US$)Limit
Enterprise subscriptionnot publishedper-customer Order Form after a demo

The site has only a demo request and a login. The terms of service say that customers pay subscription fees set in an Order Form, invoiced in US dollars and due within 30 days, and that contracts renew for 12 months at a time, with renewal price rises capped at 5% unless the customer grows a lot.

02

When did Aida launch, and how did it grow?

From two search engineers who started a sales assistant in 2024 to an acquisition by a legal AI company, in at most 33 months. Each entry carries its grade.

  1. 2024CPrahladka and Das Sarma found Aida in Palo Alto. They spend more than a year building before they bring the product to market, the company said. [4, 3]
  2. October 2025DDealroom records a seed round of US$11M. [3]
  3. 15 July 2026CPress release on the launch of the Digital Twin. Customers include Eve Legal, WorkWave, Synctera, Straiker and TeamOhana, and usage is above 80%, the company says. [4]
  4. 29 October 2026AHarvey announces the acquisition, its fifth of 2026. No price is disclosed. The team joins Harvey's product and engineering organization. [1, 2]
  5. November 2026Cgetaida.com is still live, with the line "Aida is joining Harvey" and a link to the announcement. [5, 8]
03

Why did Aida work?

Four causes, in the order they mattered. Each one is a reusable lesson.

01

One place for the facts of a deal

A rep's knowledge of a deal sits in email, calendar, call recordings, Slack and the CRM. Aida connects to all of them (Salesforce, HubSpot, Gmail, Zoom, Slack and Gong among others) and answers questions such as "What did we quote them?" from one store.

02

Act, not only record

The company set itself against conversation intelligence tools, which record and analyze calls. Aida drafts the follow-up, updates the deal record, flags risks and prepares the next meeting brief, and the rep approves the action.

03

Design partners before the price

The founders pitched sellers and let them describe the product, Prahladka wrote. Eve Legal worked with Aida as a design partner for seven months before it became a customer, and its feedback shaped pipeline views, email sync and custom prompts.

04

A team a buyer wanted

Harvey did not buy a legal product. Its announcement says the team's skill in search and machine learning will help each law firm build its own ontology, a structured map of the firm's internal knowledge.

How it was built

These details come from Aida's own pages and its press release (grade C), and from the page source of its website (grade B). [9, 8, 10, 4, 3]

  • Connectors to Salesforce, HubSpot, Gmail, Zoom, Slack and Gong. The company says setup takes 30 minutes and that reps keep their usual tools.
  • Several ways in: a web app, a Chrome extension for the inbox, a Slack app, a desktop recorder for in-person and online meetings, a mobile app, and an opt-in text-message service for registered users.
  • Support for the Model Context Protocol (MCP), an open standard for connecting AI applications to outside tools and data, so customers can add sources without a custom integration.
  • SOC 2 Type II certification, a security audit standard that business buyers ask for. The platform page says Aida never trains models on customers' calls, deal data or rep activity.
  • No downloaded source names the language model provider.
  • The marketing site is built with Framer and measured with PostHog, according to its page source.

Tools named in the sources

Each one opens its page in the Tools tab, with every story that uses it.

04

Who tried this before Aida, and why did they fail?

Three companies in our database of failed startups promised to turn the scattered data of salespeople into insight or action, and failed. The values below are from that database and are not independently verified.

Died 2012 to 2016Burned US$2M

Brisk

The promise

Turn salespeople's data from CRMs, emails and calendars into insights that help them sell.

Why it died

Each client needed custom integration work, which used up resources faster than revenue came in.

Unit Economics
Died 2017 to 2020Burned US$1.2M

Slik

The promise

On-demand sales intelligence that gathers and analyzes data for sales teams.

Why it died

It could not stand apart from LinkedIn Sales Navigator and ZoomInfo, and could not keep its data fresh.

Competition
Died 2021 to 2024Burned US$15M

Toplyne

The promise

Predict which free users of a software product will pay, and send them to the sales team's tools.

Why it died

Each customer needed custom data work, sales budgets shrank in the layoffs, and analytics platforms added the same scoring.

No Market Need

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

Brisk tried in 2012 what Aida did in 2024: collect a salesperson's CRM, email and calendar data in one place. It needed custom work for each client. Slik and Toplyne sold insight to sales teams, and larger platforms offered the same insight inside products the teams already paid for.

Two things differ. Language models can read unstructured text, such as emails and call transcripts, without rules written for each customer, and standards such as MCP cut the work of each new connection. And Aida did not have to win the market alone: Harvey bought the team for its search and machine learning skill, which Harvey wants to apply to law firms' own knowledge.

05

Can one person copy Aida now?

The product idea, yes. The path, not for one person.

Aida was two experienced founders with venture money and a team that Dealroom puts at 39, selling to enterprise sales teams through signed order forms. One person cannot copy that sales path. The part that carries over is narrower: connect the places where one kind of worker keeps facts, and let the model do the follow-up work instead of only reporting it.

Where it is still open: The same assistant for roles with smaller software budgets than enterprise sales: recruiters, account managers at agencies, and owners of small service businesses.

  1. 1Pick one role whose facts are spread over several systems, as a sales rep's are over email, calls and the CRM.
  2. 2Work with a few design partners for months before charging, as Aida did with Eve Legal.
  3. 3Make the assistant act (draft, update, flag) and let the user approve each action.
  4. 4Pass the security audits that business buyers ask for before they connect their email and calls.

What could go wrong for Aida?

The risks are the mirror image of the causes above.

What is not verified

No revenue figure was found, and neither Harvey nor Aida disclosed the price. No payment-processor profile, founder interview with figures or filing was found. The usage figures (above 80% weekly active usage, 67% faster sales cycles, 2.5 times quota capacity, three times the output of one rep in beta) are the company's own claims. The headcount of 39 comes only from Dealroom's directory, read on 8 November 2026, after the sale, and the founding month in 2024 is not published. Dealroom's funding record (a US$11M seed round in October 2025 from Ascolta Ventures) does not name the investors that Aida names (8VC, CRV and Firebolt), so it may be incomplete. Harvey does not say whether Aida stays on sale to sales teams. A reader could look for a founder interview or Harvey's later product notes.

Meeting recording and consent

Aida's desktop recorder "Records any meeting, anywhere — in-person, web, desktop", and the Eve Legal case study says that Aida "doesn't appear as a visible note-taking bot during meetings". Some laws require the consent of every participant before a conversation is recorded. The terms of service place this on the customer, who "represents and warrants that it has or will obtain prior to using the Platform any authorizations and consents required". No downloaded source says how the other people in a meeting are told about the recording.

A Digital Twin of each rep

The Digital Twin learns a rep's writing style from that rep's sent emails and call transcripts and drafts emails that "sound like you"; in the home page example the rep approves and sends the draft. The company says each twin is built only from the rep's own data, that the rep can review and reset it, and that a new rep can start from a top performer's patterns "with permission". It writes text drafts for the user's own email, not voice or video.

Competition and model providers

Eve Legal's leadership also looked at Gong and Kaya, according to Aida's case study, and Aida's press release positions it against the established conversation intelligence tools. On this site, Klipy sells a similar AI assistant that keeps a sales team's CRM up to date, from a much smaller team. No downloaded source names the language model provider behind Aida; the company page says only that "LLMs finally make it possible" to build it, so its quality and cost depend on outside models that the sources do not identify.

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.

  1. A Harvey blog, 29 October 2026: Aida Joins Harvey
    Read in full. The buyer's own announcement. Source for the acquisition, the founders' names and backgrounds, the team joining, the fifth acquisition of 2026 and Harvey's reasons.
  2. B Dealroom News, 29 October 2026: Harvey snaps up Aida in its fifth acquisition of 2026
    Read in full (downloaded 7 and 8 November 2026). An independent news summary. Source for the undisclosed sum.
  3. D Dealroom company profile: AIDA
    Read the public profile, 8 November 2026. A directory record. Source for the founding year, the 39 employees, the team growth and the US$11M seed round of October 2025.
  4. C Aida press release, 15 July 2026: Aida Launches AI Digital Twin Platform
    Read in full. The company's own release. Source for the founding year, the named customers, the usage claims, the investors and the build time before launch.
  5. C Aida company page: open letter by Piyush Prahladka
    Read in full. Source for the founder's background, the origin of the idea, the investors, the offices and the line on joining Harvey.
  6. C Aida case study: How Eve Legal's Founding AE Managed 200+ Accounts with Aida
    Read in full. Source for the seven-month design partnership, the customer's account counts, the recording without a visible bot and the competitors the customer looked at.
  7. C Aida terms of service
    Read the sections on the Order Form, consents, fees and term. Source for the paid subscription and the customer's duty to obtain consents.
  8. C Aida home page
    Read in full, 8 November 2026. Source for the product surfaces, the usage claims, the Digital Twin example and the page source (Framer, PostHog).
  9. C Aida platform page
    Read in full. Source for the integrations, the 30-minute setup, SOC 2 Type II and the rule against training models on customer data.
  10. C Aida: Introducing Digital Twin in Aida, and Aida Now Supports Model Context Protocol (MCP)
    Read both posts in full (the MCP post at getaida.com/resources/aida-now-supports-model-context-protocol-(mcp)-connect-your-ai-sales-assistant-to-more-tools-and-data-sources). Source for what the Digital Twin learns, its privacy controls and MCP support.
  11. C Database of failed startups, local export
    Local 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

Back to all stories