The headline number is not what makes Udaipur-based Edysor.ai interesting. The voice-AI startup has raised just Rs 1.2 crore in a pre-seed round led by angel investor Jasmine Sarupria — a modest cheque by any measure, and one you could easily scroll past in a daily funding roundup. What deserves a second look is how the company charges: not for tokens, not for compute minutes, but for outcomes. Edysor bills clients for results, not runtime.
That pricing choice puts a small Indian startup in the middle of one of the more consequential debates in enterprise AI right now. As buyers grow weary of open-ended usage bills that scale with consumption rather than value, outcome-based pricing is emerging as a plausible answer — and voice AI, where a “result” is unusually easy to define, may be where the model gets tested first. Here is a look at the bet.
The raise and model
According to Entrackr, which first reported the deal, Edysor.ai raised Rs 1.2 crore in a pre-seed round led by angel investor Jasmine Sarupria. The figure was corroborated the same week by StartupTalky’s July 8 funding roundup. The company says it will use the fresh capital to accelerate product development, sharpen its AI capabilities, grow the team and expand its customer base.
Edysor is not a brand-new venture. It was founded in January 2020 by Atishaya Jain, Abhinav Jain, Gunjan Pancholi and Fivos Radis, and pivoted into AI in 2024. Today it builds voice-AI agents aimed at enterprise lead conversion — the machine on the other end of the line that qualifies prospects, follows up and pushes deals along a pipeline.
The differentiator is the meter. Rather than charging by usage — per minute, per token, per seat — Edysor operates on an outcome-based pricing model, billing customers based on the results its agents deliver rather than the runtime they consume. In a category where most vendors price on consumption, that is a deliberate contrarian stance, and it is the reason a sub-crore pre-seed is worth writing about at all.
Why outcome pricing matters
Enterprise AI has spent the last two years running on usage-based billing. You pay for tokens processed, minutes streamed, API calls made. The logic is clean for the vendor — costs scale with compute — but it puts the risk squarely on the buyer. If the AI runs, you pay, whether or not the run produced anything you can bank.
Outcome-based pricing inverts that. When a vendor charges for results, it takes on the delivery risk and aligns its revenue with the value the customer actually receives. For an ROI-focused buyer — a sales head who cares about qualified leads, not voice minutes — that alignment is the entire pitch. You are no longer paying for an AI to try; you are paying when it works.
The catch, and it is a real one, is definition. Outcome pricing only functions where the outcome is measurable, attributable and hard to game. Voice AI for sales qualification is a comparatively good fit: a qualified lead, a booked meeting or a converted call is a discrete, countable event. That is likely no accident. The sectors where results are legible are exactly the ones where this model can be underwritten. In fuzzier applications — brand sentiment, creative assistance, general copiloting — nailing down what counts as a “result” is far harder, which is why usage billing still dominates there.
Edysor is not alone in reaching for this frame; outcome- and results-based pricing has become a talking point across the voice-AI space and enterprise software more broadly. But talking about it and building a business on it are different things. The interesting question is whether the unit economics hold once the invoices are tied to outcomes the vendor does not fully control.
Early traction (and caveats)
The early numbers Edysor reports are encouraging, with the emphasis on early. Per StartupTalky, the company has signed 15 customers in two months since going live, reached roughly $3,000 in monthly recurring revenue, and seen voice-minute consumption climb about 300% month-on-month. Entrackr adds that Edysor is bringing on two to three new customers a week and running between 50,000 and 100,000 AI-powered calls per month, with the company claiming 15-to-20x ROI on its AI-powered sales qualification.
Those deployments span a wide swathe of the economy — real estate, BPO, D2C, automotive, edtech, healthcare and BFSI — with named clients reportedly including Chanakya University, Ambuja and Global Tree, among others.
Now the caveats, because they matter. These are the numbers of a company two months into a product launch. A $3,000 MRR is a signal of pull, not a proof of a business; 300% growth off a tiny base is arithmetic as much as momentum; and a 15-to-20x ROI figure is the company’s own claim, not an independently audited result. Fifteen logos in eight weeks says the pitch resonates, but retention, expansion and payback over a full year are the tests that separate a promising demo from a durable model. On outcome pricing specifically, the thing to watch is gross margin: if Edysor is carrying compute cost while only billing on results, a run of low-conversion campaigns could squeeze it. None of this is disqualifying. It is simply the difference between traction and proof.
The India read
Read against the Indian market, the bet gets sharper. India is a voice-first, multilingual, cost-sensitive market where a great deal of business still happens over the phone — in real estate, education, lending and insurance especially. Voice AI that can operate across Indian languages and dialects is not a novelty here; it is a fit for how a large part of the country actually transacts. That is fertile ground for agents that qualify leads and chase follow-ups at a fraction of a call-center’s cost.
Results-based pricing sharpens the wedge. Indian enterprise buyers are famously ROI-focused and wary of open-ended software bills. Telling a sales head “pay us when we deliver a qualified lead” is a far easier conversation than asking them to forecast voice-minute consumption for a technology they have not yet trusted. It lowers the barrier to the first yes — and in a market where the first yes is the hard part, that is a genuine go-to-market advantage.
Where does voice AI win first? Most likely in exactly the high-volume, outbound, clearly-measurable use cases Edysor is targeting — sales qualification, appointment-setting, collections follow-ups — before it moves to subtler, higher-stakes conversations. Whether Edysor specifically becomes the company that captures that opportunity is unknowable from a pre-seed round. But the direction of travel it is betting on — voice interfaces for Indian enterprises, priced on outcomes rather than runtime — is a reasonable read of where a meaningful slice of the market is heading. As enterprises tire of paying for tokens, the vendors who can credibly charge for results will have a story the buyers are ready to hear.
