{"slug": "dhravya-shah-raises-3m-for-supermemory", "title": "Dhravya Shah Raises $3M for Supermemory", "summary": "Dhravya Shah raised approximately $3 million in seed funding for Supermemory, a startup building a universal memory API for AI applications. The round was led by Susa Ventures, Browder Capital, and SF1.vc, with backing from Google AI chief Jeff Dean and Cloudflare CTO Dane Knecht. Supermemory extracts structured insights from unstructured data to provide persistent context for large language model applications.", "body_md": "### What happened\n\nPer TechCrunch (Ivan Mehta, Oct 6 2025), **Dhravya Shah** raised a **$2.6 million** seed for **Supermemory**, a startup building a universal memory API for AI applications. TechCrunch is the most authoritative single source for the deal terms; multiple Indian outlets and the founder's own X post describe the round as $3 million (including a possible post-close top-up), so the exact final figure is in mild dispute - accounts differ by $0.4M. The round was led by **Susa Ventures**, **Browder Capital**, and **SF1.vc**, with individual investors including Google AI chief **Jeff Dean**, Cloudflare CTO **Dane Knecht**, DeepMind product manager **Logan Kilpatrick**, and Sentry founder **David Cramer** (TechCrunch; Yahoo Finance; ASU entrepreneurship).\n\n### What Supermemory does\n\nSupermemory extracts \"memories\" or structured insights from unstructured data - files, docs, chats, emails, app data streams - and builds a knowledge graph to surface relevant context for applications (TechCrunch). Supported integrations include Google Drive, OneDrive, and Notion. Early customers at announcement included **Cluely** (a16z-backed), AI video editor **Montra**, AI search **Scira**, and Composio's Rube tool (TechCrunch). The startup competes with Letta (Felicis-backed), Mem0, and Samsung-backed Memories.ai; Shah claims differentiation through lower retrieval latency (TechCrunch).\n\n### Context - why the story is circulating now\n\nThe original funding was announced in October 2025. Indian media, including Times of India (June 27, 2026), republished the story in an educational \"what students can learn\" format - not a new funding event. For practitioners, the underlying infrastructure problem (persistent multi-session memory for LLM agents) remains an active unsolved space.\n\n### Practitioner takeaway\n\nMemory-layer tooling sits at a genuine engineering gap: models still lose context across sessions, and retrieval-augmented generation alone does not fully address multi-session coherence. Watching latency benchmarks, privacy/access-control implementation, and customer expansion beyond the initial cohort will indicate whether Supermemory differentiates from its competitors in production use.\n\n## Key Points\n\n- 1Supermemory raised ~$3M seed (TechCrunch: $2.6M; founder's X post: $3M) backed by Jeff Dean, Dane Knecht, Susa Ventures, and Browder Capital in October 2025.\n- 2The startup provides a universal memory API for LLM apps - extracting structured knowledge graphs from unstructured data to enable persistent cross-session context.\n- 3Watch latency/cost benchmarks and access-control design: those will determine whether Supermemory's knowledge-graph approach outperforms RAG alternatives in production.\n\n## Scoring Rationale\n\nA niche but genuinely useful AI infrastructure seed from October 2025, re-surfaced in June 2026 Indian educational media. The memory-layer problem is real and the backers are credible, but this is seed-stage, 8-month-old news. Score reflects solid relevance to AI practitioners without overstating the significance of a ~$3M early round.\n\nPractice interview problems based on real data\n\n1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.\n\n[Try 250 free problems](/problems)", "url": "https://wpnews.pro/news/dhravya-shah-raises-3m-for-supermemory", "canonical_source": "https://letsdatascience.com/news/dhravya-shah-raises-3m-for-supermemory-9b29cd09", "published_at": "2026-06-27 05:28:30+00:00", "updated_at": "2026-06-27 06:38:23.336850+00:00", "lang": "en", "topics": ["ai-infrastructure", "ai-tools", "ai-startups", "large-language-models", "ai-agents"], "entities": ["Dhravya Shah", "Supermemory", "Susa Ventures", "Browder Capital", "SF1.vc", "Jeff Dean", "Dane Knecht", "TechCrunch"], "alternates": {"html": "https://wpnews.pro/news/dhravya-shah-raises-3m-for-supermemory", "markdown": "https://wpnews.pro/news/dhravya-shah-raises-3m-for-supermemory.md", "text": "https://wpnews.pro/news/dhravya-shah-raises-3m-for-supermemory.txt", "jsonld": "https://wpnews.pro/news/dhravya-shah-raises-3m-for-supermemory.jsonld"}}