Cogent Labs vs Mihup

Neither company has publicly disclosed a valuation, making a definitive size comparison difficult. Cogent Labs employs 100-500 people, while Mihup has 50-200 employees.

Cogent Labs has raised more in total funding at $50M, compared to Mihup's $6M — a gap of $44M. Combined, the two companies have completed 6 known funding rounds.

Cogent Labs leads with an Awaira Score of 55/100, while Mihup sits at 50/100. That 5-point gap reflects real differences in funding, scale, and traction — it's not a vanity metric.

Cogent Labs was founded by Andrew Hall in 2016. Mihup was founded by Biplab Kumar Jha in 2016. Visit each company's profile on Awaira for a full founder biography.

Cogent Labs: Cogent Labs develops AI document processing and text analysis software with specialised capabilities in Japanese character recognition, handwritten text, and structured document data extraction, targeting the Japanese financial services, insurance, and government sectors where large volumes of handwritten and mixed-format documents require processing with high accuracy in Japanese scripts including kanji, hiragana, and katakana. The Tokyo company applies deep learning models trained on large Japanese document datasets to achieve recognition accuracy on complex Japanese text that general OCR systems cannot match.\n\nThe company raised approximately $50 million in venture funding from investors including WiL and Goldman Sachs. Cogent Labs Tegaki handwriting recognition product has been deployed by major Japanese insurance companies, financial institutions, and public sector organisations to automate document digitisation workflows that previously required large manual data entry teams. The Japanese market for document AI is substantial given the volume of paper-based documents in government and financial services operations and the complexity of Japanese script that makes off-the-shelf OCR insufficient.\n\nCogent Labs competes in the Japanese document AI market against OBIC, NTT Data, and international intelligent document processing vendors including ABBYY and Kofax. Its Japanese-specific technical capabilities create a natural market advantage that English-first vendors struggle to match through localisation alone, as accurate Japanese handwriting recognition requires specialised model training that cannot be derived from models built primarily on Latin character datasets. The company has expanded internationally to address Korean and other East Asian scripts with similar character recognition complexity. Mihup: Mihup is a conversation intelligence platform specializing in speech recognition and NLP for Indian languages, offering real-time call analytics, voice search, and speech-to-text capabilities tuned for Indic accents and multilingual code-switching common in Indian call centers. The platform serves BFSI, telecom, and retail customers that operate large vernacular customer support operations.\n\nThe company raised approximately $6M in Series A funding and has built proprietary speech models for Bengali, Hindi, and other regional languages that outperform global ASR providers on Indian accent benchmarks. Mihup's products are deployed in production contact centers processing high daily call volumes.\n\nAccurate speech AI for Indian languages remains an underserved technical problem given the phonetic diversity and code-switching patterns of Indian speakers. Mihup's focus on this specific challenge, combined with years of proprietary training data collected from real call center environments, gives the company a technical moat that is difficult for English-first global vendors to replicate quickly. Mihup competes in the Indian conversation intelligence space against Enthu.ai, Convin, and global platforms like Gong and Chorus that lack deep Indic language support. The company's core advantage lies in its proprietary speech recognition models trained specifically on Indian language datasets, delivering higher accuracy on Hindi, Tamil, Bengali, and other regional languages than English-first competitors can achieve. Mihup has gained traction in BPO, banking, and insurance sectors where call center operations in regional languages represent billions of minutes of untapped analytics potential across the Indian subcontinent.

Both Cogent Labs and Mihup launched in 2016. Same year, but even a few months' head start matters in AI — early movers lock in data, talent, and customer relationships fast.

Cogent Labs has about 100-500 employees; Mihup has about 50-200. A bigger team usually means more revenue or heavier VC backing, but in AI, small teams can build at massive scale.

Yes — they're direct rivals. Both Cogent Labs and Mihup compete in NLP, targeting many of the same buyers. If you're evaluating one, you should be looking at the other.