AI Startups To Watch: 5 Startups That Caught Our Eye In November

Startups & Funding

India's AI Money Chases Deployment, Not Frontier Models: Five Startups to Watch

AuxoAI's $20.7 Mn raise and Blacksmith's $66 Mn unicorn round anchor a shift: India's AI capital now targets deployment, routing and agents over frontier models.

By
Rebecca Stone
Filed
Channel
Startups & Funding
Read
5 min read

AuxoAI raised $20.7 million in late October, and Blacksmith pulled in $66 million at a unicorn valuation — the clearest signal yet that India's AI investors are betting on companies that deploy existing models, not the next frontier lab. QwikBuild added $2.3 million for its enterprise AI platform, and Spill Games secured $10.5 million for AI-led mobile gaming.

For two years, the loudest question in Indian boardrooms was which large language model an enterprise would standardise on. That question has largely been settled. The challenge has shifted to running these models efficiently, reliably and at scale — a problem especially acute for enterprises running multiple models, where a model strong enough for complex reasoning may be too expensive for routine queries.

The near-term prize, by this logic, looks less like a fresh frontier model and more like the enormous market for serving existing ones cheaply and reliably. Inc42's 14th edition of its Five AI Startups To Watch series, a curated alphabetical selection rather than a ranking, captures that shift across inference routing, semantic video intelligence, autonomous commerce, AI-native CRM and pipeline generation.

Who made the November list?

  • BLUE — AI-native video data lake
  • Kily — autonomous ecommerce agents
  • Revspot — AI revenue infrastructure
  • RouterPlus — LLM routing gateway
  • Superleap — AI-native CRM

BLUE: Turning CCTV into a searchable layer

Most enterprise CCTV footage goes unwatched until something breaks, and searching it means sifting through months of data. BLUE, founded in 2023 in Bengaluru and led by CEO Kunal Kislay alongside Ashish Mehta, Mohit Garg and Nishant Sinha, compresses meaning at capture rather than scanning raw pixels downstream.

Its codec cuts storage by up to 90% and compute at the source by up to eight times, encoding objects, events and behaviours as structured data. Teams can query their cameras in plain English, with agents tuned for retail, warehousing, oil and gas and manufacturing. BLUE runs cloud, on-premises or hybrid. It counts Britannia and Hindustan Petroleum among customers in a global video analytics market projected to reach $33.61 billion by 2031.

Kily: Agents that run the marketplace playbook

A consumer brand with a few hundred SKUs can face thousands of campaigns and price checks across Amazon, Flipkart, Myntra, Blinkit, Zepto and Instamart — more than any human team can track. Kily, founded in March 2025 by ex-Flipkart executives Sankalp Mehrotra and Anurag Singh with ex-Amazon Ads executive Sharad Chitlangia, deploys an autonomous agent that configures, monitors and optimises itself rather than recommending actions.

Its Core Decision AI coordinates specialist agents that set bids and budgets, watch price parity and flag stock-out risks. Kily positions itself as a P&L management layer for ecommerce and claims to work with over 50 brands, including ITC, Unilever, Mondelez and Coca-Cola, in India's retail ecommerce market, which Redseer projects will hit $90 billion in 2026.

Revspot: Pipeline generation for high-ticket sales

In real estate, education and wealth management, stale customer data, delayed follow-ups and low-intent leads quietly destroy conversions. Revspot, founded in 2024 by Darshan Subash, Chirag Wadhera and Varun Garg, builds an AI growth agent called Spot that decides who to target, what to say and where to reach them — then launches campaigns, enriches leads, runs voice agents and follows up over WhatsApp. Nothing goes live until a human approves.

The Bengaluru startup says it serves more than 50 customers across real estate, education, automotive and wealth management, and raised $4.8 million in a Series A co-led by Inflexor Ventures and Info Edge Ventures, with Pentathlon Ventures, Silicon Road Ventures and Titan Capital. It operates in a global AI sales agent market estimated at $3.4 billion in 2026 and projected to reach $15.8 billion by 2033.

RouterPlus: One key for smart LLM routing

Teams building on LLMs now juggle multiple providers, API keys and contracts, often picking models on global benchmarks that say little about their own workloads. RouterPlus, founded in 2026 by Alok Bishoyi and Lakshmi Narayana, routes each request based on a customer's own evals and production traffic rather than leaderboards.

Its catalogue spans closed and open-weight models — Claude, GPT, Gemini, Qwen, DeepSeek — and the gateway offers reserved capacity through providers such as Modal, Fireworks and Baseten. It stores no prompts, supports regional routing and bring-your-own-key, and helps teams shift from closed to open-source models. The AI model router market it targets is estimated at $100 million in 2025 and projected to reach $4.1 billion by 2035.

Superleap: Rebuilding CRM as a revenue OS

Sales reps feed legacy CRMs with data but rarely benefit, so they skip updates and the systems rot. Superleap, founded in 2024 by former Unacademy operators Subham Kumar Boundia, Alok Maurya and Saurabh Maheshwari, is rebuilding the category as an AI-native revenue operating system.

The stack comprises SuperOS, a unified data layer; SuperSense, a customer-intelligence layer; and SuperAgents, an agentic layer automating tasks across the revenue cycle, plus AI voice bots and MCP integrations with WhatsApp, Slack and ChatGPT. Its Superleap Deploy team promises migration off legacy CRMs in under two months with zero disruption. Razorpay, Aakash Education, Cars24 and MediBuddy are customers in a global CRM market projected to reach roughly $321 billion by 2034, at a 12.4% CAGR.

The pattern across all five is consistent: investors and founders alike are pricing the plumbing, autonomy and measurement around existing models — and that market, not frontier research, is where India's near-term AI returns will likely be decided.

Source: inc42.com

Share this article:

More from Rebecca Stone

Rebecca Stone

Show full bio

Correspondent covering technology and product news in Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.

1,037 articles

Related articles

« Previous articleNext article »