Program Lead - Data Platform Program

The role

Cloudsufi is hiring a Program Lead to own, end to end, a multi-year data platform program for a strategic technology client: a public knowledge-graph and statistical data platform that ingests, refreshes and publishes thousands of datasets, plus a portfolio of custom deployments for enterprise, government and research organisations worldwide. The program is delivered by a ~25-person engineering team across the US and India. The client's senior sponsor expects a partner who owns the program, not one who executes tickets: reports that land right the first time, financials that are never late, work that is visible in the client's own tools, and issues resolved before the sponsor hears about them. You will be the single point of accountability for all of that, and the technical authority the client turns to when the hard questions come up.

This is a hands-on leadership role. You will read the code reviews, challenge the architecture, understand why a pipeline failed, and set the quality bar yourself. You will also run the commercials, the governance and the executive relationship. We are looking for someone who has done both at once, for a demanding client, and would rather be measured on outcomes than on activity.

What you will own:

• Program ownership: Single point of accountability to the client sponsor and the client's core engineering team. Issues escalated to the sponsor that we should have resolved ourselves: zero.
• Technical leadership: Set and enforce the quality bar before anything reaches the client: complete PRs with issue links and validation evidence, a recurring-issue log driven to zero, first-pass review acceptance rising quarter on quarter.
• Delivery throughput: Agree and freeze scope each iteration; deliver at least 90% of committed scope; roughly double dataset import/refresh throughput within two quarters, with cycle time by size tracked against a baseline.
• Executive reporting: Monthly newsletters and the quarterly business review accepted by the sponsor on first draft; pre-reads with our CEO one week out and with the sponsor five days out; one clear message per update; every document opens with a review-and-approval table.
• Financial discipline: Live financials updated by the 2nd of each month, figures agreed by the 5th, invoices closed with evidence by the 5th, every approval on email. Budget burn tracked against the SOW with no surprises.
• Work visibility: 100% of team work logged daily in the client's tracking tools and per-phase time tracking; nothing reaches the client without your review.
• Team structure: A lean, unambiguous org: one owner per role, measurable success criteria for every role, a RACI the client can read in one pass; weekly reviews with a current decision log.
• Shaping, not following: Grow our share of contributions to the client's ingestion and refresh process (tools, agents, validation checks, runbooks), tracked in a contribution register co-signed by the client.
• Commercial stewardship: Own SOW amendments, scope-versus-contract alignment and the scaling model for new custom deployments, with the Client Partner and Finance.

Hands-on experience:

We need:

• You do not need to have done every item, but you must be able to sit with an engineer, open the code and get to the root cause yourself.
• 15+ years in data engineering, platform engineering or technical program leadership, with 5+ years leading programs of $3M+ a year for a demanding enterprise or hyperscaler client.
• Large-scale data ingestion and refresh pipelines: schema design and validation, statistical and time-series data at hundreds of datasets and tens of millions of data points, data provenance and licensing compliance, source qualification, failure triage and root-cause analysis.
• Google Cloud at depth: BigQuery, Spanner, GCS, Cloud Run / GKE, Vertex AI (including embedding models and regional deployment), Looker. GCP Professional Data Engineer or Architect certification preferred.
• Agentic and LLM-assisted engineering: you have put code-review assistants, diagnostic agents or skills frameworks into a real delivery workflow, managed hallucination and prompt-quality issues, and measured the velocity gain. MCP and agent integration into client applications is a strong plus.
• Custom deployments in customer-hosted environments: cloud infrastructure, networking and security reviews, deployment runbooks, embedded UI (iframe, D3 / web components) and conversational agents bolted onto a product.
• Contributing to a client's platform: PRs that improve the product itself (cost analysis, validation, configuration-driven frameworks), reviewed and merged by the client's core team.
• Distributed teams: led 20+ engineers across US and India time zones, with execution leads reporting to you and a technical lead co-located with the client's core team.
• Program governance: iteration planning and scope freezes, RAID, QBRs, SOW and PO management, fixed-fee and T&M billing, velocity and penalty clauses, financial forecasting to the dollar.

How you work:

• You own outcomes, not tasks. When something goes wrong, your first instinct is "this is mine to fix," not "this is blocked on the client." You resolve problems before the sponsor hears about them and tell them what you did, not what you need.
• You get it right the first time. You would rather send one clear page than three drafts. You review every number before it leaves the team, and you treat a report that needs rework as a defect.
• You are hands-on by choice. You read the PRs, you can explain why the pipeline failed, and you set the technical bar by example. Engineers respect you because you can do the work, not because of your title.
• You are relentless about visibility. If work is not in the system, it did not happen. You enforce daily updates, per-phase time tracking and decision logs because you believe in them, not because the client asked.
• You communicate like the client's own leadership. One message per update, value and asks up front, risks stated plainly, a review-and-approval table on every document. You adapt to how the client's organisation writes and decides.
• You are financially disciplined. Invoices, forecasts and approvals close on the date, with evidence, every month. You treat late billing as a trust problem, not an admin problem.
• You shape rather than follow. You bring proposals to the client's core team, get them adopted, and move the partnership from "executes well" to "co-designs the process."
• You build lean, clear teams. One owner per role, measurable criteria for every person, no ambiguity about who decides. You give credit by name and hold people to the standard the best of them already meet.
• You are calm under pressure. You have been on a program in code-yellow, turned it around in weeks, and kept the team's confidence while doing it.

Qualifications and logistics:

• Bachelor's or Master's in Computer Science, Engineering or a related field, or equivalent experience.
• Based in the San Francisco Bay Area with the ability to be on-site with the client several days a week.
• Authorised to work in the United States.
• Start date: ASAP.

First 90 days:

Take over program governance and reporting, own the next quarterly business review, and deliver the first measurable throughput gain.

Skills

BigQuery, Data Engineering, GCP, GCS, GKE, Google Cloud, Looker, Platform Engineering, Program Management, Spanner