Headline

Builder | AI-First Founder | Founding Engineer → Director of Engineering @Atlan | 600+ K8s Clusters · $3M Cloud · 100+ Engineers | MokshVed — AI Infra on K8s

About

Most infrastructure looks fine right up until the day it has to hold. I spent 11 years across SocialCops and Atlan — joining as a founding engineer and ending as Director of Engineering running systems that made enterprise-grade reliability possible for a data governance platform with no business being as reliable as it was. The numbers: 600+ Kubernetes clusters across AWS, Azure, and GCP. 4–5 PB of observability data on VictoriaMetrics and ClickHouse. $3M+ in annual cloud savings — not from cutting tools, from building the architecture right the first time. 100+ engineers hired, including the VP and Director pipeline that now runs multiple departments. Zero Trust, SOC2, and ISO 27001 built from scratch. That infrastructure supported Tier 1 banks, media companies, and automotive firms across 4 continents. Built while reporting directly to the Founder & CEO — owning the cloud budget, partnering with enterprise GTM and CX, and making architectural calls that had to be right before anyone ever noticed them. The culture side mattered as much. Recognised by both Co-Founders as Values Champion — not for writing the values, but for living them: giving direct feedback without ego, letting engineers own their work fully, and building an environment where people grew without needing to be managed. The teams I built were reliable because they trusted each other, not because someone was watching. None of that work demoed well. Most of it was invisible until the day it held. Right now I'm building MokshVed — an AI orchestration layer on Kubernetes for teams that need AI to run on infrastructure they own, not infrastructure rented from a third-party API. Live work. If you're scaling infrastructure, hardening a platform for the first time, or building serious AI workloads and want to trade notes — I'm open to that conversation.

Location

New Delhi

AI stack

codex

cursor

claudecode

Work experience

Founder at MokshVed, Dec 2025 - Present · 9 mos, India, Building MokshVed — an autonomous AI orchestration layer on Kubernetes. Most teams scaling AI workloads hit the same wall: the infrastructure that was fast to spin up becomes the bottleneck that's impossible to own. They end up renting complexity from managed services instead of building systems they control. MokshVed replaces the manual coordination layer with intelligent, self-optimising platform agents — so engineering teams can run serious AI workloads on infrastructure they own, without the infrastructure tax eating their time and margin. Live work.

Director - Engineering at Atlan, Apr 2023 - Oct 2025 · 2 yrs 7 mos, India, Joined Atlan as a founding engineer. Left as Director of Engineering. The work in between is what mattered. The infrastructure: 600+ Kubernetes clusters across AWS, Azure, and GCP at 99.99% uptime. 4–5 PB of observability on VictoriaMetrics and ClickHouse. $3M+ in annual cloud savings — from building the architecture right the first time, not cutting tools. Zero Trust network, SOC2, and ISO 27001 built from scratch. All of it supporting Tier 1 banks, media companies, and automotive firms across 4 continents. The org: 100+ engineers hired. 4–5 EMs running core platform teams. The VP and Director pipeline I built now runs multiple departments. Reported directly to the Founder & CEO — owned the cloud budget, partnered with enterprise GTM and CX, made architectural decisions that had to be right before anyone noticed them. The philosophy: let engineers own their work fully. Give direct feedback without ego. Build an environment where people grow into roles they didn't think they were ready for. Former direct reports now hold Director, and Staff Engineering titles globally. Recognized by both Co-Founders as Values Champion — not just for living the values, but for defining them.

Director - Cloud/Security/IT at Atlan, Oct 2020 - Apr 2023 · 2 yrs 7 mos, New Delhi, Delhi, India, Three engineers. No playbook. Had to go from startup architecture to enterprise-grade before the enterprise customers arrived. The landmark: onboarded Hindustan Unilever (HUL) — India's largest FMCG — as Atlan's first major enterprise customer. Passing their security and reliability bar from scratch. That engagement became the template for every enterprise deal that followed. The architecture: built the multi-cloud backbone across GKE, EKS, and AKS as a core team of three, supporting three separate product offerings while scaling to 100+ enterprise customers. Provider-agnostic from the start — so customer expansion never required infrastructure rework. The security foundation: led Atlan's first SOC2 push, building security into the development lifecycle before it was a compliance requirement. That work became the basis for Zero Trust and ISO 27001 implementation as the company scaled.

Functional Lead - Cloud & Security at Atlan, Mar 2019 - Sep 2020 · 1 yr 7 mos, New Delhi, Delhi, India, The beginning. Before there was a security posture, a deployment standard, or an enterprise customer — someone had to make the foundational calls. Built Atlan's first cloud security architecture from scratch: identity management, network isolation, and encryption protocols. Not as a compliance exercise — as the actual technical foundation the company would build on for years. Standardised the engineering workflow with Helm. Set up Terraform for infrastructure-as-code from day one. The underlying principle: if it wasn't in code, it didn't exist. That decision held through 600+ clusters and years of scale. Piloted the core cloud-native toolchain and CI/CD patterns that became Atlan's deployment culture — the defaults that 100+ engineers eventually inherited.

Engineering Lead at SocialCops, Mar 2018 - Mar 2019 · 1 yr 1 mo, Came back to engineering and led it. Owned technical delivery, introduced Terraform and infrastructure-as-code practices, and established the engineering standards that followed me into the Atlan infrastructure build.

Project Manager at SocialCops, Sep 2017 - Sep 2018 · 1 yr 1 mo, Led delivery of national-scale government data projects including Disha and Ujwala — coordinating across central government stakeholders, field teams, and engineering under hard public-sector deadlines.

Product Manager at SocialCops, Sep 2016 - Mar 2018 · 1 yr 7 mos, Moved from building the product to owning it. Defined roadmap and requirements for SocialCops' data intelligence platform — translating government agency and development organisation needs into product decisions.

Engineering Fellow at Jaaga Study, Jan 2014 - Aug 2014 · 8 mos, Bengaluru, Karnataka, India

Information Technology Consultant at Freelance, Sep 2012 - Aug 2014 · 2 yrs

Education

Bachelor of Technology - BTech, Electronics and Communication Engineering, Panipat Institute of Engineering and Technology

Dr. MKK Arya Model School