Headline

Founder, KyraDB | Context infra for reliable engineering agents | MPL

About

Founder - Kyra DB. Kyra, a self-improving enterprise context & memory layer for AI agents. Every enterprise has years of data scattered across Salesforce, Jira, Slack, and databases - messy, siloed, and meaningless to agents. KyraDB connects to all of it, understands what it actually means, and builds a persistent causal knowledge graph that compounds over time. Agents stop starting cold. Unlike RAG, KyraDB understands why things happened, not just what was said. The more it is used, the smarter it gets. With experience spanning early-stage ventures and software development, I have architected innovative solutions, including decision-grade memory systems and multi-model AI runtimes, to improve reliability and operational efficiency. My focus remains on creating robust, scalable solutions to help enterprises navigate complexity and achieve operational excellence.

Location

Bengaluru

AI stack

claude code

claude

codex

cursor

deepseek

hugging face

Work experience

Founder at Kyra, Feb 2026 - Sep 2026 · 8 mos, Bengaluru, Karnataka, India, Kyra, a self-improving enterprise context & memory layer for AI agents. Every enterprise has years of data scattered across Salesforce, Jira, Slack, and databases - messy, siloed, and meaningless to agents. KyraDB connects to all of it, understands what it actually means, and builds a persistent causal knowledge graph that compounds over time. Agents stop starting cold. Unlike RAG, KyraDB understands why things happened, not just what was said. The more it is used, the smarter it gets.

AI Solutions Consultant at Self Employed, Apr 2025 - Dec 2025 · 9 mos, Bengaluru, Karnataka, India, Architected a decision-grade memory system treating recall as a constraint on reasoning — improving consistency, reducing drift, and enabling convergence in autonomous agents. • Designed selective recall and causal eligibility frameworks for decision-relevant retrieval • Built contradiction guardrails and an immutable fact layer to stabilize cross-session intelligence • Reduced token usage and latency under identical workloads • Orchestrated OpenAI, Claude, and DeepSeek within a unified multi-model runtime • Developed agent-driven browser systems for autonomous retrieval and synthesis

Growth & Ops Consultant at Shiba Inu, Jul 2024 - Mar 2025 · 9 mos, Remote, - Deployed RPA pipelines to automate daily operations (scheduling, rewards), reducing manual overhead by 60%. - Integrated AI-based engagement triggers, resulting in an 18% boost in user retention. - Achieved $2,000/month recurring revenue with zero marketing spend through product-led growth hacking. - Negotiated high-impact ecosystem partnerships with Mirai Labs, AstraNova, and ReadyPlayerMe.

Education

Master of Business Administration (MBA), Operations and Marketing, Great Lakes Institute of Management

B.Tech, Chemical and Bio Engineer, Dr. B R Ambedkar National Institute of Technology, Jalandhar ( PUNJAB)