Dreamteam, a CRM startup, said in a company post dated Nov. 17 that it has raised $53 million across its seed and Series A rounds, led by Together Fund and Accel with participation from Premji Invest. The money backs a specific bet about sales software: a CRM that records what buyers say on calls, in emails and in meetings can replace one that depends on reps typing notes into fields.
What Dreamteam announced
The post is signed by co-founders Anand Venkatraman, Ramesh Parthasarathy and Senthil Kanthaswamy. It says more than 100 companies across 13 countries use the product to run their sales. It gives no split between the two rounds and no valuation. The three write that they have spent more than 25 years between them building and selling B2B software, running pipeline reviews and chasing reps for missing updates.
Their diagnosis is about data entry. “A CRM was supposed to hold that story and help the whole company work from it. Instead, it became a database that reps have to feed and managers have to police,” wrote Venkatraman, Parthasarathy and Kanthaswamy, Dreamteam’s co-founders. They argue that the signals deciding a deal, such as a prospect hesitating on price or a competitor’s name coming up on a call, often never reach the record because dropdowns and text fields have nowhere to put them.
How the product is built
Four AI teammates on one memory
Dreamteam describes four AI teammates. Rachel briefs a seller on a prospect and its company. Sally keeps deals moving and drafts follow-ups. Raj watches the pipeline and flags risks. Alex turns a business need into CRM setup; in the company’s example, a sales leader asks for a security-review stage in one sentence and the pipeline changes.
Behind them sits what the company calls the Company Brain, which keeps each customer’s story in one place and draws from the team’s calls, emails and meetings. The post says the CRM updates itself, and that when an account changes hands the new owner inherits the customer’s goals, past objections and open commitments.
Where people stay in the loop
The company puts limits on the agents in the same post. Pricing rules, approval limits and company policies act as guardrails, and “no email goes out without a person’s approval.” Users can see where a recommendation came from and correct what Dreamteam remembers. A sales team decides what to delegate and stays accountable for promises made to a customer. RevOps teams weighing those controls can start from the argument we made for setting a CRM data baseline before any sales agent pilot starts.
The counter-view from a16z
One investor’s published view sets a limit on how far a new CRM can get. In a May 14 post, Andreessen Horowitz’s Gio Ahern, Stephenie Zhang and Alex Immerman wrote that “The CRM isn’t going to go away” and that it is turning into one of many inputs to what they call systems of intelligence. They wrote that Salesforce and HubSpot “still sit on some of the most valuable datasets in the industry” and are building API-first offerings. Their own go-to-market survey, they added, found CRM usage has risen since AI tools were adopted at scale, because agents that write structured call notes back into the system have made its data richer.
The a16z authors expect most of the next decade’s value in go-to-market software to land in the reasoning layer above the database. Dreamteam is building the database and the reasoning layer as one product. In our reading, both positions put the value in conversation context, and they differ on whether a buyer has to change the system of record to capture it. Our coverage points the same way on the evidence so far: Menlo found limited evidence of AI replacing CRM, and HubSpot is leaning into a headless version of its own product.
One vendor’s switching numbers
Switching does happen in vertical markets. Unanet, which sells ERP and CRM software to government contractors and to architecture, engineering and construction firms, said on Nov. 16 that almost half of its new customers in 2026 migrated from another ERP or CRM solution. The figure is the vendor’s own and comes without a customer count.
Unanet ties the moves to data quality. Its 2026 AEC Inspire Report found that 75% of AEC firms are using AI and only 29% have high confidence in the underlying data. “AI is only as useful as the data behind it, and our customers’ project, financial and pipeline data already lives in Unanet,” said Jeff Davison, Unanet’s Chief Customer Officer.
What it means for the sales leader
A CRM replacement pitch in 2027 planning will rest on capture. The vendor will say the record fills itself from conversations, so forecast reviews stop depending on what reps typed. That claim can be tested in a pilot, and the test is cheaper than a migration. It also changes who owns the system. If Alex-style setup agents change pipeline stages from a sentence, RevOps needs a log of who asked for each change and a way to reverse it.
The incumbents’ answer will be that the same capture can run on top of the CRM a team already has. The a16z survey finding supports that view, and so does the richer data its authors describe flowing back into existing systems.
How to evaluate an AI-native CRM
- List the sources that feed the record (calls, email, meetings) and confirm the team already records and stores each one.
- Ask the vendor to correct a wrong fact in the demo and show where the correction is stored.
- Write down which actions need a person’s approval, and check that RevOps can set pricing rules and approval limits without vendor help.
- Ask for references that replaced a CRM and references that ran the product beside one. Dreamteam’s post gives a customer count and no replacement figure.
- Run 30 days on one team next to the current CRM and compare the two forecast calls at the end.
A team that finishes that list will know whether it is buying a new system of record or a capture layer for the one it has, and can budget for the right project.
Source: Dreamteam

