Startup Deep Dive : Rezo.ai — Processes 72 million calls monthly, yet started as a text automation platform struggling to gain traction

Rezo.ai processes 72 million calls monthly across 10+ languages, yet started in 2018 as a text automation platform struggling to gain traction. The IIT Delhi-founded conversational AI startup discovered its breakthrough not through additional funding, but through a single customer—Maruti Suzuki—who showed the company that its technology could unlock untapped value in voice, not just chat.

Today, Rezo.ai is EBITDA-positive and processing 3 million calls daily for enterprises across NBFSI, automotive, and telecom. The company generates ₹42.7 crore revenue in FY25 (up 27.8% from ₹33.5 crore in FY24) with just 145 employees, after peaking at 105 two years earlier during an ill-timed expansion. What most startups struggle to find—consistent unit economics and a repeatable sales playbook—Rezo solved by fixing its go-to-market strategy and picking a single customer vertical deep enough to matter.

Quick facts

What they do

Rezo.ai operates a unified platform for customer experience automation, replacing manual call center agents with AI agents that understand intent, context, and sentiment. The platform handles inbound customer support (complaint resolution, FAQ responses, appointment booking, order tracking), outbound engagement (lead qualification, reactivation campaigns, collections), and back-office operations (audit automation, data extraction).

• Voice bots handling appointment bookings, callbacks, and customer inquiries across Hindi, English, Tamil, Telugu, Kannada, and 5+ other Indian languages, plus English variants for global calls
• Chatbots and virtual assistants across WhatsApp, chat, email, and SMS—often the first touchpoint for high-volume, low-complexity queries
• Robotic process automation (RPA) for back-office tasks like ticket creation, data reconciliation, and lead assignment
• Real-time analytics and sentiment tracking showing enterprise dashboards of call outcomes, customer mood, and agent productivity
• Integration with Azure OpenAI services for natural language understanding and generation

The company’s customers span NBFSI (non-bank financial services, including lending and collections), automotive (dealers and OEMs), telecom, consumer durables, and banking. Maruti Suzuki, Delhivery, CarDekho, Spinny (used-car marketplace), and Livguard (battery maker) are among publicly named clients.

The origin

Manish Gupta and Rashi Gupta are IIT Delhi graduates—Manish in computer science, Rashi with a double master’s in mathematics and computing plus a PhD from the University of Helsinki in natural language processing and machine learning. They met during their college years and launched their first data analytics startup in 2012. After that venture, the couple identified a different problem: Indian contact centers were inefficient, understaffed, and losing customer interactions to poor agent training and high attrition.

In 2018, they founded Rezo.ai with the conviction that conversational AI—software trained to understand customer intent and respond naturally—could automate the high-volume, low-skill interactions that consumed 60–70% of contact center agent time. The insight was simple: if agents spend most of their day answering “Where is my order?” or “How do I reset my password?”, an AI trained on thousands of such conversations could handle the same query in seconds, freeing agents to handle disputes, escalations, and complex requests that actually need human judgment.

The founders spent two years building and testing the core product—initially focused on text-based chat automation for websites and mobile apps. Rashi led the data science and NLP work; Manish drove product strategy and early customer engagement. By 2020, the team had a working MVP but faced the first test: could they sell it to real enterprises?

The struggle years

From 2018 to early 2020, Rezo.ai remained a pre-revenue startup seeking product-market fit. The early product—a chatbot platform for text conversations—found interest among some Indian startups and smaller BPOs, but lacked the momentum or reference customer that would signal to large enterprises that the software worked at scale. Pivoting from concept to a paying customer base meant navigating multiple barriers: educating the market about conversational AI when most contact centers still relied on IVR systems from the 2000s, building trust in a technology that many feared would replace jobs (rather than augment them), and proving ROI in a sector obsessed with cost-per-call metrics rather than customer lifetime value.

By late 2020, the company had a small number of customers but no signature win. Flat monthly pricing ($50K–$100K per month for a typical enterprise) made unit economics fragile—a lost customer meant missed runway. The team had pivoted the business model multiple times: from pure SaaS to hybrid models involving implementation services, from selling to small e-commerce platforms to targeting larger enterprises, from emphasizing cost savings to emphasizing customer experience uplift.

In December 2020, the founders raised ₹$282K in a seed round from Modulor Capital, ThinkNEXT Technologies, Dexter Ventures, and 30 other angels including Bhavesh Manglani (Delhivery co-founder) and Devesh Sachdev (Fusion Microfinance founder). The capital was enough for 18–24 months of runway at minimal burn, but the message was clear: investors believed in the founders and the market, but the company still needed to prove the model worked.

The turning point

The breakthrough came in late 2020–early 2021 through Maruti Suzuki, the country’s largest carmaker by volume. Maruti ran approximately 90 lakh (9 million) customer interactions annually through its franchise service centers but captured only 1.5 years of service revenue per customer against an 11-year lifetime value window. The gap was simple: Maruti’s 500+ service centers were overwhelmed with inbound calls asking for appointment slots, service status, and warranty details. The company could not proactively reach customers for preventive maintenance, vehicle upsell, or loyalty offers because its teams were drowning in reactive inquiries.

Maruti and Rezo.ai began a pilot in early 2021. The critical insight: voice automation—not chat—was the unlock. Maruti’s customers were calling to book slots; they did not want to type. Rezo.ai’s founders realized that their text-focused product would not win in a market where 80% of customer interactions were still voice. They pivoted the roadmap to prioritize conversational voice AI, building agents that could answer calls naturally, take appointment bookings, and handle objections without a human agent.

Within 18 months, Maruti saw:

• 8,500 leads generated daily (up from near-zero proactive outreach)
• 6 lakh (600,000) service appointments booked monthly through automated voice calls
• 60% reduction in total operational cost for the customer service function
• 75% reduction in human agent workload per center

The Maruti deployment proved three things: (1) enterprises would pay a premium for demonstrated ROI, not flat fees; (2) voice automation was underexplored and had massive TAM (total addressable market) in India; and (3) a single reference customer in a large, respected company could unlock sales conversations across an entire vertical.

Armed with the Maruti case study, Rezo shifted its go-to-market approach in 2021–2022. Instead of selling a generic “customer service platform,” the company targeted automotive dealers and OEMs with a specific offer: “We reduced Maruti’s service cost by 60%; let’s do the same for you.” The sales narrative changed from “chat automation” to “voice automation for high-volume outbound and inbound.” Pricing shifted from flat monthly fees to value-based models where Rezo’s margin scaled with customer savings (typically 40–70% lower cost than human agents for handled calls).

The money behind it

Rezo.ai has raised $282K in disclosed funding to date, all from a single seed round on December 15, 2020. The round was led by Modulor Capital, with institutional participation from ThinkNEXT Technologies (a seed-stage VC firm) and Dexter Ventures. The syndicate also included angels:

• Bhavesh Manglani (Delhivery co-founder): Brought knowledge of logistics and supply-chain automation, plus enterprise credibility in India’s startup ecosystem.
• Devesh Sachdev (Fusion Microfinance founder): Contributed BFSI (banking, financial services, insurance) domain expertise and a network of potential customers in the lending and collections verticals.
• 31 other angels from IIT Delhi alumni networks, NASSCOM, and the startup community.

The modest seed size ($282K) reflected the founders’ bootstrapped mentality and the VCs’ belief that Rezo did not need massive capital upfront to prove the model. Instead, the focus was unit economics: if the company could generate positive gross margin on each customer deal and retain them, reinvested revenue would fund growth. By 2024, the company had reached 51 employees and $4M in annual revenue (reported ARR), without raising a Series A. As of August 2025, Rezo had ~145 employees and was targeting ₹80–85 crore ARR (~$10M USD) by end of FY26.

Notable: Rezo has not announced Series A funding despite strong revenue growth. The company is likely prioritizing profitability and free cash flow over rapid dilution, a stance consistent with India’s post-2022 venture-capital reset where late-stage valuations have compressed and venture returns have been questioned.

How it makes money

Rezo.ai operates a pure SaaS model with a hybrid pricing approach evolved over the company’s lifecycle:

• Early model (2018–2020): Flat monthly subscription ($50K–$100K per month), priced per enterprise, often bundling the platform with implementation and consulting services. This model was margin-negative on customer acquisition; churn was high because customers could not easily measure ROI against a fixed cost.
• Value-based pricing (2021 onward): Rezo shifted to charging based on demonstrated cost savings. For example, if Rezo’s system handles 10,000 calls per month and costs ₹50 each to handle manually, Rezo might charge the enterprise ₹30–35 per call (40–70% savings vs. human agent cost of ₹50), splitting the margin with the customer. This aligns incentives: Rezo profits when the customer profits.
• Volume-based pricing: Some enterprise contracts involve a base fee plus a per-call overage charge, allowing Rezo to scale revenue as customer volume grows.

Gross margins are estimated at 60–70% once the product is live (platform cost per call is ~₹5–10 including infrastructure, OpenAI API, and support; enterprise pricing is ₹30–50 per call). Customer acquisition cost (CAC) is high (~$50K–$100K for enterprise sales) but paid back within 6–12 months because average contract values (ACVs) have increased to ₹15–30 lakh (~$18K–$36K USD) annually as value-based pricing has matured.

The company operates with minimal direct sales staff—as of 2024, only 2 full-time sales representatives carried individual quotas, while the Maruti case study and industry reputation drove inbound interest. Rezo reinvests revenue into product development (NLP improvements, new language support, agentic AI capabilities) and customer success (implementation engineers, post-sale technical support).

The numbers

Key metrics as of 2024–2025:

• Call volume: 72 million calls processed monthly across all channels; 3 million calls daily; processing power across 10+ languages and 500+ dialects.
• Call value extraction rate: 92% (meaning 92% of calls result in a desired outcome—appointment booked, complaint resolved, payment made, etc.—without human escalation).
• Cost per call handled: ₹5–10 (platform and infrastructure only), priced at ₹30–50 to customer, yielding 60–70% gross margin.
• Average contract value: ₹15–30 lakh annually; shifted upward from $50K–$100K flat fees once value-based pricing took hold.
• Paying customers: 15 named references (as of Founder Thesis interview, 2024); customer base includes auto dealers, NBFCI platforms, telecom, and logistics.
• Profitability: EBITDA-positive for 8+ consecutive quarters; the company is not burning cash and is reinvesting revenue into product and headcount.

Customer lifetime value is estimated at ₹1.5–3 crore per multi-year enterprise contract, based on typical retention of 3–5 years and account expansion (landing as a voice automation vendor, expanding to chatbots, RPA, and analytics).

Where the money comes from

Rezo’s revenue is split across three primary segments (disclosed implicitly through case studies and product marketing):

• Automotive & OEM (40–50% of revenue, estimated): Maruti Suzuki, dealership networks, and 2–3 other major OEMs. These customers use Rezo primarily for service appointment booking, warranty upsell, and post-sale customer engagement. ACVs are highest in this vertical (₹50 lakh–₹1+ crore annually) because the ROI is clearest—every appointment booked is ₹10K–15K revenue for the dealer.
• NBFSI (30–40%): Collections, lending origination, and customer service for microfinance companies, fintech platforms, and traditional banks. Rezo is used for outbound collections calls, lead qualification for loans, and customer support for disbursements/repayment. These customers are highly price-sensitive but have massive call volumes (1M+ calls/day per customer is not uncommon).
• Other (logistics, telecom, e-commerce) (10–20%): Delhivery (order tracking, delivery confirmations), CarDekho (lead qualification), and smaller telecom/utility companies. Revenue here is growing but smaller per customer.

Within each segment, the “surprise” is not the revenue concentration but the margin improvement. Automotive dealerships and NBFCs typically manage 500+ customer service reps at annual cost of ₹100+ crore per company; Rezo’s platform, handling 70–80% of volume, costs ₹15–30 crore annually. For an NBFCI processing 50 lakh (5 million) collections calls monthly, Rezo’s platform is cheaper than hiring additional agents and delivers better collections rates (because the AI can work 24/7 without fatigue and can handle objections more patiently than a fatigued human agent). Rezo has published that its system achieves 10% improvement in collection efficiency (collections agents close ~30% of cases; Rezo-handled cases + escalation achieve ~33%).

The risks

• Market concentration risk: Rezo’s top customer (Maruti) likely represents 15–25% of revenue based on the scale of the deployment (600K appointments monthly × assumed fee per appointment). Losing this customer or a contract renegotiation could impact growth. The company has mitigated this by diversifying into auto dealer networks and NBFSI, but automotive is still the dominant vertical.
• Infrastructure dependence on Azure OpenAI: Rezo’s NLP engine runs on Microsoft Azure and OpenAI APIs. Any outage, price increase, or API discontinuation (e.g., if OpenAI shifts to a proprietary model licensing model) could disrupt Rezo’s service or margin. Rezo has some in-house NLP capability but is not fully vertically integrated.
• Regulatory tightening on AI and voice calling: India’s regulatory environment (TRAI, RBI, NCLAT) is increasingly focused on unsolicited calls, data privacy, and consent. Outbound voice automation is especially at risk if regulators impose stricter rules on robo-calls or require manual call review. A blanket ban on outbound voice automation would disrupt Rezo’s high-value NBFSI and auto dealer businesses.
• Commoditization of voice automation: Global players (Amazon Connect, Google Cloud Contact Center AI, Twilio Flex) are adding conversational AI capabilities, and Indian startups (e.g., Observe.AI, Callbot) are entering the space. Rezo’s moat is execution (Maruti case study, value-based selling playbook, India-native language support) rather than IP; price competition could compress margins if larger players subsidize voice automation to gain market share.
• Series A funding pressure: As growth accelerates, hiring for 145→250+ headcount will require cash. Rezo has bootstrapped this far and is EBITDA-positive, but scaling sales and product teams in a competitive market often requires external capital. If Rezo does not raise Series A, it may face a slower growth trajectory relative to well-funded competitors or may face acquisition pressure from larger enterprises seeking to own the technology.

The takeaway

Rezo.ai teaches a counterintuitive lesson: early-stage companies often over-correct on product breadth and under-correct on go-to-market strategy. Rezo built a solid text-based conversational AI platform, but the company did not achieve meaningful traction until it pivoted (1) to voice, (2) to a single vertical with extreme depth (automotive), and (3) to value-based pricing that aligned incentives with the customer. The Maruti deployment was not a lottery—it was the outcome of speaking to 50+ automotive customers, hearing the same pain (appointment scheduling was broken), and building the product to solve that pain visibly.

Most startups would have tried to be “the Salesforce for customer service”—selling to 200 different verticals with generic chat, email, and voice automation. Rezo instead became “the appointment booking system for car dealers,” then used that moat to expand to collections, then to logistics, then to e-commerce. By going deep in one vertical before spreading wide, Rezo generated a repeatable sales playbook and a reference customer base that converted venture-backed competitors into non-threats because Rezo had already proven ROI in each vertical before the VC wave landed.

Frequently asked questions

How does Rezo.ai differ from chatbot platforms like Intercom or Freshchat?

Rezo is primarily a voice automation platform for contact centers; Intercom and Freshchat focus on text (chat, email) for product support and sales. Rezo’s strength is high-volume outbound voice (e.g., 10,000 appointment reminders per day), while Intercom is strongest in inbound support for SaaS. Rezo handles voice across 10+ Indian languages; most chat platforms prioritize English. However, Rezo is expanding into multi-channel (chat + voice + email); the distinction is narrowing.

Why did Rezo not raise Series A funding if revenue is growing so fast?

Rezo is EBITDA-positive and reinvesting revenue into product and sales, reducing the need for dilutive capital. Series A is typically raised when a company needs cash to fuel growth faster than revenue allows, or when founders want liquidity. Rezo’s founders appear to be prioritizing profitability and ownership retention. As hiring scales (145→250+ headcount), the company may eventually raise a round for executive hires, geographic expansion, or to counter well-funded competitors.

What is Rezo.ai’s market size?

India has approximately 10,000+ contact centers (including BPOs, in-house centers, and micro-centers) handling ~500 million customer interactions annually. The total addressable market (TAM) for conversational AI in contact centers is estimated at $500M–$1B annually in India alone (50–100 crore calls daily at ₹5–50 per call savings). Rezo’s serviceable market is narrower: high-volume enterprises in auto, BFSI, and telecom where ROI is measurable and buyers have budget authority (~$100M–$200M). The company’s current ₹42.7 crore revenue represents <1% of even the serviceable market, indicating headroom for 5–10x growth without market saturation.

Is Rezo.ai a good model for other Indian B2B SaaS startups?

Rezo’s playbook—deep vertical focus, value-based pricing, reference customer obsession, and profitability over growth-at-all-costs—is increasingly relevant post-2022. Indian B2B SaaS startups pursuing this model (Chargebee, Razorpay, Postman) have achieved scale and profitability. Rezo shows the pattern is not a fluke; founders who resist VC pressure to “disrupt” everything and instead focus on a real customer problem can build sustainable businesses.

Will Rezo.ai go public or be acquired?

As of September 2026, there are no public signals of an IPO or acquisition. The company’s profitability and modest cash burn make it self-sufficient. An IPO is possible if growth accelerates to ₹500+ crore revenue and the founders want liquidity, but India’s current IPO market favors high-growth SaaS companies (>50% YoY), and Rezo’s 27–50% growth (FY25–FY26 estimated) is solid but not VC-grade. Acquisition is possible if a large enterprise software vendor (Salesforce, HubSpot, or an Indian player like Kissflow) sees Rezo’s voice automation and want to bundle it into a broader customer success platform.

Sources

Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics).

• Inc42, “Rezo.ai — An Enterprise Tech Funded Company Based Out Of Noida” (Company profile), 2020–2026
• Tracxn, “Rezo.ai — 2026 Company Profile, Team, Funding, Competitors & Financials” (Database), August 2025
• YourStory, “This IITian couple’s startup is leveraging ML to enhance customer experience,” 2019 (Profile + founder interview)
• Founder Thesis, “The AI powered call centre of the future | Rezo.ai,” 2024 (Founder interview with detailed business model and Maruti case study)
• Latka, “How Rezo.ai hit $4M revenue with a 51 person team in 2024,” 2024 (Financial metrics and revenue breakdown)
• Rezo.ai, “Case Studies” (Web, accessed September 2026; includes Maruti Suzuki, Livguard, Spinny, IIT Delhi results)
• Rezo.ai, “About Us” (Web, accessed September 2026; company mission, product capabilities, team details)
• Microsoft, “Rezo.ai: From contact centres to revenue centres,” 2024 (Azure AI First Movers program)
• Entrepreneur, “Rezo.ai Raises Seed Funding Led By Modulor Capital” (Press, December 2020; funding round details)
• BW Disrupt, “Rezo.ai Raises Seed Funding Led By Modulor Capital,” 2020 (Investor list and round details)
• Silicon India, “AI Startup Rezo.ai raises Seed Funding from Modulor Capital & Others,” 2020 (Investor list and background)
• MyGov Blogs, “Indian Entrepreneurs Solving Social Issues: The AatmanirbharAI Challenge Winners, 2020,” (Recognition and founder profiles)

Found an error? Write to us and we’ll correct it in the open, dated, on the piece.