Kamalakar D

Founder & CEO @ SandLogic, a deep-tech AI company building private, Edge/on-prem AI infrastructure for enterprises operating in regulated, latency-sensitive, and data-critical environments.Over time, it became clear to me that enterprise AI only scales when the entire stack is designed as a system—from compute and runtime to models, rather than as cloud-first tools stitched together. SandLogic is built around this systems-first approach.The ProblemAs enterprises adopt AI across language, speech, and edge systems, cloud-native and black-box solutions struggle with data privacy, compliance, cost, and latency. In many real-world environments, AI must run on-prem or at the edge, with predictable behavior and full ownership.What We Are BuildingSandLogic is building a full-stack AI infrastructure, including:ExSLerate AI Co-Processor: an ultra-low-power AI accelerator designed for real-time inference in constrained and privacy-sensitive environmentsEdgeMatrix Runtime: a portability and acceleration layer enabling efficient AI execution across CPUs, GPUs, and edge devicesFoundation Models (Language & Speech):- Shakti LLM Series (100M–8B+): optimized for reasoning, multilinguality, and on-prem / edge deployment- Sruthi (ASR): multilingual, noise-robust speech recognition for real-world enterprise audio- Svara (TTS): natural, low-latency speech synthesis for embedded and enterprise deploymentsTogether, these components enable enterprises, OEMs, and system integrators to deploy end-to-end AI systems - from speech and language understanding to edge inference, fully on-prem or in controlled environments.Why SandLogicSandLogic is engineered for enterprises that require data sovereignty, predictable latency and cost, edge-to-enterprise scalability, and production-grade AI infrastructure.We are built with global enterprise deployment in mind, including regulated customers across India and the US.Recognition- Aegis Graham Bell Award (Government of India–backed)- National Startup Award- Forbes Select 200 – Globalist- 1st Runner-Up, Maruti Suzuki Incubation Cohort- Selected under India’s Chips-to-Startup (C2S) program for indigenous AI chip developmentEcosystem & Strategic Programs- Selected under AWS ML Elevate / Generative AI cohort- Selected for Qualcomm Startup Mentorship Program (QSMP) focused on edge AI and on-device inference