Go back

Affordable AI Will Be Built in India First: Redrob CEO Felix Kim Makes the Case to Hundreds of Indian VCs

Authors:

Felix and Redrob Research Team

Date:

May 8, 2026

At the Korea-India Emerging Market Delegation in Gurugram – held alongside President Lee's state visit – Kim argued that the world's cheapest LLM infrastructure won't come from Silicon Valley. It will come from a market where 63 million businesses have zero AI and every token costs 5x too much.

NEW YORK / GURUGRAM - April 21, 2026

————————————————————————————————————

Felix Kim, CEO of Redrob Inc., presenting "Affordable LLMs for Emerging Markets" at the Korea-India Emerging Market Delegation Amphitheatre at The Circle FC, Gurugram, April 21, 2026.

Redrob Inc. (redrob.io) today announced that its CEO, Felix Kim, presented to hundreds of Indian venture capital firms at the Korea-India Emerging Market Delegation, held at the Amphitheatre at The Circle FC in Gurugram on April 21, 2026. The event was hosted by Korea's Ministry of SMEs and Startups and the Korea Institute of Startup & Entrepreneurship Development (KISED) as part of a three-day delegation running alongside Korean President Lee Jae Myoung's state visit to India.

India First. Then the World.

Kim presenting Redrob's market expansion roadmap: India now, Southeast Asia by 2027, US mid-market by 2028.

Kim opened with a number the audience knew but had never framed this way: 63 million small and medium businesses in India have virtually zero AI integration. Not because they chose not to adopt it. Because they are priced out. Every frontier LLM - GPT-4, Claude, Gemini - charges per token, and Indian languages cost 4.5-8x more tokens than English for identical content. A Hindi chatbot costs a mid-market Indian company $262,000-$365,000 more annually than an identical English one. India has 1.4 billion people, 600 million under 25, and 44 million enrolled college students. The cost barrier is not a footnote. It is the reason the world's most populous country is also the world's most AI-underserved.

The solution, Kim argued, will not come from Silicon Valley - because Silicon Valley has no incentive to solve it. Frontier AI companies optimize for the highest-paying customers, not the largest underserved markets. Redrob's approach inverts this: Llama-Redrob (LoRA-finetuned Llama 3 8B/70B/405B and Llama 4 Maverick) optimizes tokenization for 22 Indian languages at a fraction of frontier cost. FLUX image generation adds visual AI. The ensemble runs on AWS Bedrock with on-device inference for the smallest model. Prove it in India's 1.4 billion-person market, then scale to every economy priced out of frontier AI: Southeast Asia by 2027, the US mid-market (6 million companies under 500 employees, most with zero AI) by 2028.

Why Indian VCs Are Paying Attention

India's AI market is projected to reach $17 billion by 2027. The government has committed $1.25 billion to its IndiaAI Mission. Indian VC is actively searching for AI infrastructure plays that are not wrappers around OpenAI's API. Yet India's most prominent AI startups - Sarvam AI, Krutrim - are building foundation models from scratch, requiring hundreds of millions in compute and years of development. Redrob's approach is structurally different: take the best open-source models, adapt them via LoRA at 200x less cost, and build the application layer (HR, GTM, Passport, EOR) that turns inference into revenue today - not after the next fundraise.

The presentation was part of the three-day Korea-India Emerging Market Delegation (April 20-22), which ran alongside President Lee's summit with Prime Minister Modi. One day earlier, Kim delivered the keynote at the Korea-India Talent Exchange & Startup Fair at The Leela Palace in New Delhi, addressing hundreds of Indian professors and students.

Redrob has raised $14 million from Korean institutional investors including Korea Investment Partners, KB Investment, Kiwoom Investment, KDB Capital, DS&Partners, Murex Partners, Daekyo Investment, and Wanted Lab. The company partners with 500 Indian universities, departments, and colleges, has 3M+ users, and counts Robostar (LG Electronics) and Hanwha Vision (Hanwha Group) among its enterprise customers.

About Redrob

Redrob

(redrob.io) is an AI research company building proprietary large language models and AI-powered applications for emerging markets. Headquartered in New York, with teams across the US, South Korea, and India.

Redrob LLM -

Proprietary 5-Model Ensemble on AWS Bedrock, including Llama-Redrob (LoRA-finetuned for 22 Indian languages) and FLUX image generation. On-device AI, 1M-token context, deep research, real-time web retrieval. The Android of LLMs.