Rezolve Ai and Tech Mahindra Forge a Global Alliance: Agentic Commerce Enters Its SI-Procured Era
Rezolve Ai and Tech Mahindra announced a global agentic commerce alliance on August 28, 2026. Here is how it differs from the TCS deal three months earlier, what buying through a systems integrator actually looks like, and why skepticism about Rezolve still matters.
Key Takeaways
- Rezolve Ai and Tech Mahindra announced a global strategic alliance on August 28, 2026, to deploy agentic commerce across retail, consumer goods and financial services at production scale
- Where the TCS deal three months earlier was a reseller agreement, this one is structured as a single path from design to live deployment, with broader industry and geographic coverage
- Deal value and revenue sharing remain undisclosed, and short sellers continue to dispute Rezolve's numbers, so buyers going through an SI need to nail down where implementation accountability sits before signing
An AI commerce vendor pairs with a 146,000-person integrator
Rezolve Ai and Tech Mahindra announced a global strategic alliance to accelerate the deployment of agentic commerce across large enterprises worldwide.
rezolve.comThe announcement landed simultaneously in New York and Pune on August 28, 2026. Rezolve Ai (NASDAQ: RZLV), which builds agentic commerce infrastructure, and Tech Mahindra (NSE: TECHM), one of India's large IT services firms, unveiled a global strategic alliance. Agentic commerce here means the pattern where an AI agent finds products, compares them and completes payment on the shopper's behalf.
The division of labor is stated plainly. Rezolve Ai supplies the commerce intelligence and transaction layer through its Brain Suite, while Tech Mahindra contributes industry expertise, enterprise integration, cloud and data capabilities, and global delivery. Tech Mahindra serves more than 1,100 clients across 90 countries with over 146,000 professionals. The target sectors are retail, consumer goods and financial services, alongside other transaction-intensive industries.
What stands out in the release is the negative framing. It explicitly commits to moving AI "beyond pilots and customer-service chatbots." Harshul Asnani, President and Head of Europe Business at Tech Mahindra, added that enterprises are ready to move from AI that answers questions to AI that completes real work. Read the other way, that is the supply side admitting how many enterprise AI commerce projects had been stalling short of that line.
How this differs from the TCS deal
This is not Rezolve Ai's first alliance with an Indian integrator. In May 2026 the company partnered with Tata Consultancy Services, and we examined that deal in detail. Two integrators in three months can look like the same story told twice. Line up the deal structures and coverage, though, and the character differs.
| Dimension | TCS partnership (May 2026) | Tech Mahindra alliance (August 2026) |
|---|---|---|
| Deal structure | Formal reseller agreement for brainpowa | Joint alliance covering design, development and deployment (reseller terms undisclosed) |
| Target industries | Primarily enterprise retail | Retail, consumer goods, financial services and other transaction-intensive industries |
| Partner scale | 56 countries, 194 service delivery centers | 90 countries, 1,100+ clients, 146,000+ professionals |
| What Rezolve brings | The brainpowa platform | The full Brain Suite (Brain Commerce / Brain Checkout / brainpowa / TraceWare) |
| What the SI brings | Resale and implementation, hands-on evaluation via Pace Port | Industry expertise, enterprise integration, cloud and data, global delivery |
| Framing of the announcement | Entry into the retail AI commerce market | Moving past pilots and customer-service chatbots |
The decisive difference is industry breadth. The TCS partnership was framed around enterprise retail. This one names consumer goods manufacturers and financial services alongside retail. Adding consumer goods matters, because brands that historically depended on retailer shelves now have an implementation partner for being discovered directly by AI agents. Extending to financial services suggests demand on that side for Brain Checkout's payment handling and for TraceWare's auditability.
The second difference is contractual form. The TCS arrangement was a formal reseller agreement to sell brainpowa into retail clients worldwide, with a clear seller and a clear product. The Tech Mahindra release says only that the two will collaborate on design, development and deployment, and says nothing about reseller terms. Deal value, revenue sharing, target pipeline and start dates are all undisclosed. That leaves more room to build things jointly, and leaves the split of responsibility to be negotiated deal by deal.
That the same vendor signed two integrators in one quarter is also evidence that neither deal is exclusive. Rezolve is not locked to a single channel and can reach both client bases. For buyers, though, it means the same brainpowa can now be purchased through two windows on different terms. You can read that as easier comparison shopping, or as a risk that implementation quality will not be consistent.
The numbers behind naming pilots as the problem
The reason vendors themselves started saying "beyond pilots" becomes clear from the research.
In June 2025 Gartner predicted that over 40% of agentic AI projects will be canceled by the end of 2027, citing escalating costs, unclear business value and inadequate risk controls. The same release flags "agent washing," the practice of rebranding existing assistants, RPA and chatbots as agents without substantive change. Gartner estimates that only around 130 of the thousands of agentic AI vendors are real.
Against that backdrop, the technical section of the Rezolve release reads differently. What it foregrounds is not model intelligence but verifiability. brainpowa is described as a family of commerce-tuned models engineered to minimize hallucination, meaning plausible-sounding fabrications, and to preserve data integrity in high-volume environments. TraceWare and Auditable AI sit on top, adding transparency and accountability across agentic workflows.
An agent that only recommends products can apologize when it is wrong. An agent that completes payment has to survive audit and dispute resolution, which requires a record of who confirmed which order on what basis. The product lineup reflects a view that the bottleneck for production deployment is the audit trail, not recommendation accuracy.
What buying through an integrator gets you, and what it misses
From the buyer's side, the practical way to read this alliance is as a procurement route.
Contract directly with a vendor and the wiring into your existing commerce platform, inventory system, CRM and payment gateway stays your problem. That is where a great many projects stall. Going through an integrator puts use-case design, integration and production deployment on one contractual line. For a company that already has core systems maintenance with Tech Mahindra, opening that path costs one additional engagement.
Two gaps are easy to miss.
The first is where implementation accountability sits. If an agent confirms the wrong order, is it the model, the integration design, or the source data? With three parties involved, isolating the cause is hard, and a loosely written SLA turns it into a standoff. Auditability features like TraceWare are better understood as the machinery that makes that isolation technically possible. It is worth specifying log retention, disclosure scope and first-line triage responsibility at contract time.
The second is that this alliance is not the integrator's only bet. One day earlier, on August 27, Tech Mahindra announced an Agentic AI Factory built jointly with Uniphore, also starting in retail and financial services, with the retail work aimed at strategic sourcing and supplier intelligence. The company's 2025 "AI Delivered Right" strategy explicitly describes a curated AI partner ecosystem spanning hyperscalers and startups. Assembling per-deal combinations is the design philosophy.
So from the integrator's vantage point, Rezolve is one strong option rather than the center of the stack. If you receive a proposal, ask why Rezolve was selected for your engagement and what it was compared against. A proposal that cannot answer that concretely is probably shaped by partner economics.
The skepticism aimed at Rezolve Ai
Judging by the announcement alone would be a mistake. Rezolve Ai attracts skepticism roughly proportional to its growth figures.
The numbers are striking. Per its SEC filing, first-half 2026 revenue came in at approximately $127 million against $6.32 million a year earlier, close to 20x, and about 2.7 times its entire audited FY2025 revenue of $46.8 million. The company reaffirmed $360 million in full-year guidance and targets an exit ARR of at least $500 million for 2026, supported by more than 1,000 enterprise customers.
Short sellers dispute this directly. In September 2025, Fuzzy Panda Research published a report alleging that Rezolve is faking ARR by acquiring declining AI startups, that its proprietary model is effectively a ChatGPT wrapper without durable advantage, and that its Microsoft and Google relationships amount to paying the platforms rather than collaborating with them. Grizzly Research separately argued that most of the revenue traces to the acquired GroupBy business. Rezolve rejected the claims as misleading and inaccurate, saying they disregard audited financials and SEC filings.
We cannot adjudicate that here. What matters is the structural point: acquisition-driven revenue and organic adoption remain hard to tell apart. That is precisely why the integrator alliances carry weight. Named production deployments through TCS or Tech Mahindra would be evidence acquisitions cannot manufacture. If announcements keep accumulating without customer names, the skeptics gain ground instead.
The market reacted calmly. RZLV rose about 1% on the day of the announcement. The stock did jump more than 20% over that week, but The Motley Fool's analysis attributes the move mainly to the August 25 news that Google selected Rezolve's distributed database platform. The Tech Mahindra alliance is still being priced as an unproven promise.
What retail and commerce operators should do now
This alliance will not change anyone's procurement tomorrow. Three points are still worth holding onto.
The first effect is simply that more options can be raised through existing integrator relationships. Evaluating agentic commerce used to come down to assembling general-purpose LLMs in house or waiting for an add-on from your commerce platform. If you already work with a global integrator on core systems, ask whether they can propose an engagement that covers agent-initiated purchasing end to end. The specificity of the answer is itself a measure of that firm's maturity.
Second, settle your RFP criteria in advance. It is telling that Rezolve evaluates its own models on commerce metrics such as sales closing rate and recommendation timing, because a pilot that measures conversation quality alone never connects to a business decision. Given how heavily this alliance leans on auditability, add how far agent decision logs can be captured, retained and disclosed to your evaluation criteria. It pays off later, in returns and mis-order disputes.
Third, prepare your own data. Whichever integrator and platform you choose, an agent cannot compare products correctly if the product information is not structured. Attribute granularity in the catalog, real-time accuracy of inventory and pricing, and how discontinued items and substitutes are expressed. None of that work depends on how this alliance turns out, so there is no reason to wait.
Conclusion
The Rezolve Ai and Tech Mahindra alliance marks the point where agentic commerce shifts from a question of whether it can be built to a question of how it gets procured. A technology vendor pairing with a global delivery organization has now happened twice, and the target industries have widened from retail into consumer goods and financial services.
What to watch over the next twelve months is how many named customers appear. The company's investor day at Nasdaq on October 6 is the first checkpoint for that. Judge this by the number of production deployments rather than the number of announcements, and the wave of press releases becomes much easier to navigate.
