August 2017. Gap Inc.’s Director of Cloud Infrastructure resigned, leaving four teams — IaaS, Private Cloud, PaaS, and Tooling — leaderless, demoralized, and behind schedule on a critical migration. Two engineers had already decided to leave. Thirteen weeks to Black Friday.
For a $15B retailer, PEAK weekend carried outsized operational and revenue risk. Previous PEAK weekends had produced outages and lost sales. The new private cloud infrastructure needed to be ready. It wasn’t. The team building it had stopped functioning.
The traditional VMware/DevOps teams in Pleasanton viewed private cloud as an existential threat to their jobs. At the leadership level there was no communication — only distance, physical and organizational.
Year one. The Director’s four teams were already demoralized — he’d been disconnected before he departed. I managed all four teams directly. I visited Pleasanton to meet each infrastructure leader one-on-one: introduced myself, heard their concerns, asked how I could help. I got to know each remote engineer individually, pushed for raises and promotions that earned trust I had no title to claim.
One senior architect, loyal to the departed director, opposed me from day one. I earned the trust of his close friend — a lead on another team under me — so thoroughly that the friend eventually told the architect to back off or lose their friendship. The private cloud platform was production-ready for PEAK, though the org chose not to risk the traffic cutover that close to the director’s departure.
Year two. A separate director saw what had happened in year one and handed me his problem — the observability team. This was the team that had stonewalled my cloud teams the year before (they once sent me a document that was just a blank template). They needed to migrate 130+ ecommerce teams from Grafana and open source tools to New Relic and Splunk before PEAK. They weren’t delivering.
I spent Wednesdays in Pleasanton doing one-on-ones with every engineer. Early on the lead engineer was openly rude to me. I went to his desk ready to swipe the headphones off his ears — and stopped myself. He complained to the director, who asked me to apologize. That evening I saw him leaving the building and introduced myself to his wife. I apologized on the way down the stairs. At the bottom he stopped, put his hand out, and thanked me. He became my strongest ally. The 130-team migration was completed before PEAK.
PEAK weekend 2017: zero downtime. Zero revenue loss.
The traditional infrastructure teams that had initially viewed the cloud teams as a threat became trusted partners. That shift was the precondition for year two.
PEAK weekend 2018: 100% of observability running on the new platforms. The CTO and CIO had full dashboard visibility across the entire event — something that had never existed at Gap before. All 130+ ecommerce teams had live dashboards to catch issues before they became incidents.
A director handed me a problem he couldn’t solve because of what I’d done the year before. That’s what earned trust looks like.
“I was very fortunate to have Seemant as my first product hire who spanned across four distinct spaces — cloud infra, application infra, build & deploy tooling, and telemetry. Seemant was trusted by his teams as well as stakeholders to lead the org in an entirely new product direction that culminated in empowering dev teams to operate much more independently and delivering solutions that significantly improved dev velocity. I genuinely miss working with Seemant, and would work with him again in a heartbeat.”
Kartik Garg · Head of Product & Tech, AI Studios — Amazon · Kartik was Seemant’s client at Gap Inc.