{ "generated_at": "2026-10-22T17:33:56.980Z", "month": "2026-10", "source": "https://saahil.co.uk/status.json", "operator": { "name": "Saahil Mehta", "title": "Associate Director, Data & Insights", "employer": "South Pole", "location": "London, UK", "url": "https://saahil.co.uk", "github": "https://github.com/saahil-mehta", "linkedin": "https://www.linkedin.com/in/saahil-mehta", "email": "self@saahil.co.uk", "book_a_call": "https://calendly.com/saahil_mehta/30min", "portrait": "https://saahil.co.uk/portrait-still.webp", "resume_pdf": "https://saahil.co.uk/Saahil-Mehta-Resume.pdf" }, "narrative": { "statement": "I lead data and AI at South Pole, and I still ship the systems myself.", "statement_key": "still ship the systems myself", "role_line": "Associate Director, Data & Insights · South Pole · London", "now": [ "I lead a team of six and own the data platform, the marts and the reporting stack for the whole company.", "Most of my time goes into Orca, the enterprise AI assistant I built for the business, and the governed platform underneath it: a semantic layer that tells Orca what our data means, self-hosted orchestration, access governance that runs itself, and cost monitoring across every project we have.", "I wrote the data strategy the executive committee adopted, and I commit code most days." ], "ledes": { "work": "The systems in production at South Pole, designed and built by me with the team.", "shipping": "Commits over the last twelve months, live from GitHub.", "experience": "Seven years across South Pole, Publicis Groupe and WPP.", "beyond": "Mentoring, a museum archive, and where I studied.", "contact": "About building AI systems people actually use, leading data teams, or your data and cloud estate.", "platform": "The systems in production, what each one does and what flows between them.", "plan": "Seven workstreams to 2028 and where each one stands." }, "end_card": "Let's talk." }, "featured": [ "Orca", "Orca Semantics", "Orca Visuals", "Data Marts Core" ], "banner": { "heading": "All systems operational", "operational": 6, "in_build": 2, "people_served": 501 }, "components": [ { "name": "Orca", "venue": "South Pole · Production", "category": "AI assistant", "status": "live", "since": "2026-06", "until": null, "months_in_production": 5, "duration": "5 months", "metric": { "value": "560,000+", "label": "requests at 99.99% success" }, "summary": "South Pole's enterprise AI assistant, open to the whole company. Deep research over the web and company data, South Pole-format documents, decks and sales proposals, scheduled runs, memory, and sharing with version history. Eight releases in four months. Within three weeks of 2.0, 131 employees were querying governed data through it; 560,000+ requests served at 99.99% success.", "opening": "South Pole's enterprise AI assistant, open to the whole company. Deep research over the web and company data, South Pole-format documents, decks and sales proposals, scheduled runs, memory, and sharing with version history.", "tags": [ "Enterprise AI", "Deep Research", "RAG", "Cloud Run", "IAM" ], "link": null, "still": "https://saahil.co.uk/stills/orca.webp", "still_light": "https://saahil.co.uk/stills/orca-light.webp" }, { "name": "Orca Semantics", "venue": "South Pole · In Build", "category": "Semantic layer", "status": "in_progress", "since": "2026-07", "until": null, "months_in_production": 4, "duration": "4 months", "metric": { "value": "6", "label": "domains, owners named" }, "summary": "A governed semantic layer on self-hosted OpenMetadata: plain-English definitions, certified metric SQL, a glossary and named owners per domain, all version controlled. Orca reads it before answering, the visualisation layer builds on it, and access approvals and quality routing reuse the same ownership model.", "opening": "A governed semantic layer on self-hosted OpenMetadata: plain-English definitions, certified metric SQL, a glossary and named owners per domain, all version controlled.", "tags": [ "OpenMetadata", "Semantic Layer", "Governance", "Terraform" ], "link": null, "still": "https://saahil.co.uk/stills/orca-semantics.webp", "still_light": "https://saahil.co.uk/stills/orca-semantics-light.webp" }, { "name": "Orca Visuals", "venue": "South Pole · In build", "category": "Visualisation", "status": "in_progress", "since": "2026-08", "until": null, "months_in_production": 3, "duration": "3 months", "metric": { "value": "~200k EUR/yr", "label": "in licences removed as Power BI retires" }, "summary": "The visualisation layer on top of Orca Semantics: dashboards and metrics defined as code against the certified definitions, so a chart and Orca's answer agree. It replaces Power BI, 98 reports and 410 users, and removes ~200k EUR a year in licences.", "opening": "The visualisation layer on top of Orca Semantics: dashboards and metrics defined as code against the certified definitions, so a chart and Orca's answer agree.", "tags": [ "Lightdash", "Semantic layer", "BI-as-code", "Power BI" ], "link": null, "still": "https://saahil.co.uk/stills/orca-visuals.webp", "still_light": "https://saahil.co.uk/stills/orca-visuals-light.webp" }, { "name": "Data Marts Core", "venue": "South Pole · Production", "category": "Orchestration", "status": "live", "since": "2026-09", "until": null, "months_in_production": 2, "duration": "2 months", "metric": { "value": "1,800+", "label": "dbt models orchestrated" }, "summary": "Orchestration for 1,800+ dbt models, migrated from dbt Cloud to self-hosted Dagster on GKE in three weeks. Shadow runs proved the output identical to production before cutover; deploys rebuild the image and roll out without a human step. Avoided a three-year renewal at ~27k EUR a year.", "opening": "Orchestration for 1,800+ dbt models, migrated from dbt Cloud to self-hosted Dagster on GKE in three weeks.", "tags": [ "Dagster", "dbt", "GKE", "Cloud Build" ], "link": null, "still": "https://saahil.co.uk/stills/data-marts-core.webp", "still_light": "https://saahil.co.uk/stills/data-marts-core-light.webp" }, { "name": "Data Foundation Plan", "venue": "South Pole · Adopted by ExCo", "category": "Strategy", "status": "live", "since": "2026-08", "until": null, "months_in_production": 3, "duration": "3 months", "metric": { "value": "~295k EUR/yr", "label": "direct savings in the plan" }, "summary": "The company data strategy, written, costed and taken to ExCo: seven workstreams to 2028 covering the semantic layer, orchestration, visualisation, the AI platform, data quality at scale, access governance and SLAs. ~295k EUR a year of direct savings, with the dbt Cloud saving already banked before the plan opened.", "opening": "The company data strategy, written, costed and taken to ExCo: seven workstreams to 2028 covering the semantic layer, orchestration, visualisation, the AI platform, data quality at scale, access governance and SLAs.", "tags": [ "Strategy", "Roadmap", "Governance", "FinOps" ], "link": null, "still": "https://saahil.co.uk/stills/data-foundation-plan.webp", "still_light": "https://saahil.co.uk/stills/data-foundation-plan-light.webp" }, { "name": "Data Access Governance", "venue": "South Pole · Production", "category": "Governance", "status": "live", "since": "2026-10", "until": null, "months_in_production": 1, "duration": "1 month", "metric": { "value": "40+ h", "label": "returned to the team each month" }, "summary": "Self-service data access. A request names the access group and the reason, mails the data owner, and opens a Terraform pull request that grants the group read access, through an org-owned GitHub App. A human still merges. Replaced hand-edited IAM at around ten changes a month and returned 40+ hours a month to the team.", "opening": "Self-service data access. A request names the access group and the reason, mails the data owner, and opens a Terraform pull request that grants the group read access, through an org-owned GitHub App.", "tags": [ "IAM", "Terraform", "GitHub App", "Governance" ], "link": null, "still": "https://saahil.co.uk/stills/data-access-governance.webp", "still_light": "https://saahil.co.uk/stills/data-access-governance-light.webp" }, { "name": "Cost Deep Dive", "venue": "South Pole · Production", "category": "FinOps", "status": "live", "since": "2025-01", "until": null, "months_in_production": 22, "duration": "1 year 10 mo", "metric": { "value": "195", "label": "GCP projects watched" }, "summary": "FinOps across 195 GCP projects and 60 services: spend by app, by service, by day, with P99 anomaly detection that emails me when an app breaks its own baseline. Caught a 6x BigQuery analysis spike in March 2026 and returned it to baseline; marts query spend fell 46% year on year.", "opening": "FinOps across 195 GCP projects and 60 services: spend by app, by service, by day, with P99 anomaly detection that emails me when an app breaks its own baseline.", "tags": [ "FinOps", "BigQuery", "Looker Studio", "Anomaly Detection" ], "link": null, "still": "https://saahil.co.uk/stills/cost-deep-dive.webp", "still_light": "https://saahil.co.uk/stills/cost-deep-dive-light.webp" }, { "name": "Enterprise Data Platform Migration", "venue": "South Pole · Production", "category": "Platform", "status": "live", "since": "2024-03", "until": null, "months_in_production": 32, "duration": "2 years 8 mo", "metric": { "value": "18,000+", "label": "resources under Terraform" }, "summary": "Migrated 18,000+ cloud resources from Pulumi to modular Terraform. Standardised IAM, networking, and policy-as-code so five business units now share one platform without stepping on each other. Adopted BI-as-code, DataOps, and FinOps practices alongside the migration, so the platform shipped governed, observable, and within budget from day one.", "opening": "Migrated 18,000+ cloud resources from Pulumi to modular Terraform. Standardised IAM, networking, and policy-as-code so five business units now share one platform without stepping on each other.", "tags": [ "Terraform", "GCP", "Infrastructure-as-Code", "CI/CD" ], "link": null, "still": "https://saahil.co.uk/stills/enterprise-data-platform-migration.webp", "still_light": "https://saahil.co.uk/stills/enterprise-data-platform-migration-light.webp" }, { "name": "Knowsee", "venue": "Private AI System for Teams · Personal Project · Paused", "category": "Side project", "status": "paused", "since": "2024-01", "until": "2026-03", "months_in_production": 27, "duration": "2 years 3 mo", "metric": { "value": "Paused", "label": "ideas folded into Orca" }, "summary": "A RAG-based AI platform that queries your data warehouse, ingests company knowledge (docs, screenshots, emails) with chunking and metadata that survives real-world inputs, and turns chat prompts into dashboards. SSO and RBAC out of the box; hallucination guardrails and sandboxed tools on by default; eval suites gate every release. Paused in 2026; the ideas went into Orca.", "opening": "A RAG-based AI platform that queries your data warehouse, ingests company knowledge (docs, screenshots, emails) with chunking and metadata that survives real-world inputs, and turns chat prompts into dashboards.", "tags": [ "RAG", "Agent Frameworks", "Vector Stores", "MCP" ], "link": "https://knowsee.co.uk", "still": "https://saahil.co.uk/stills/knowsee.webp", "still_light": "https://saahil.co.uk/stills/knowsee-light.webp" } ], "releases": [ { "tag": "v2026.07", "title": "Associate Director, Data & Insights", "org": "South Pole", "unit": null, "start": "2026-07", "end": "present", "location": "London / Remote", "summary": "Lead a team of six and own the data platform, marts and reporting stack for 501 people. Built Orca, the company's enterprise AI assistant, and the governed platform under it: a semantic layer on OpenMetadata, self-hosted Dagster, automated access governance and FinOps across 195 GCP projects. Wrote the Data Foundation Plan adopted by ExCo, ~295k EUR a year of direct savings.", "line": "Lead a team of six and own the data platform, marts and reporting stack for the whole company. Set the data strategy the executive committee adopted, and build the systems above with the team." }, { "tag": "v2024.03", "title": "Lead Specialist, Data Engineering", "org": "South Pole", "unit": null, "start": "2024-03", "end": "2026-07", "location": "London / Remote", "summary": "Owned the climate-finance data estate. Migrated 18k+ cloud resources from Pulumi to modular Terraform, shipped governed FastAPI/FastMCP services that let LLM/RAG workloads reach sensitive carbon-credit data without breaking access controls, and launched Orca 1.0 to the company in June 2026.", "line": "Own the climate-finance data estate: the move to Terraform, governed services that let AI workloads reach sensitive carbon-credit data without breaking access controls, and the first company-wide release of the AI assistant." }, { "tag": "v2023.07", "title": "Senior Data Engineer II", "org": "Publicis Groupe", "unit": "Digitas UK", "start": "2023-07", "end": "2024-03", "location": "London", "summary": "Cut flagship GCP + Airflow workloads from four hours to two. Built GitHub Actions CI/CD that replaced manual deployments.", "line": "Cut flagship GCP + Airflow workloads from four hours to two. Built GitHub Actions CI/CD that replaced manual deployments." }, { "tag": "v2022.10", "title": "Senior Data Engineer I", "org": "Publicis Groupe", "unit": "Digitas UK", "start": "2022-10", "end": "2023-07", "location": "London", "summary": "Normalised cross-channel marketing data for clients including Samsung and Formula 1, on tight campaign deadlines.", "line": "Normalised cross-channel marketing data for clients including Samsung and Formula 1, on tight campaign deadlines." }, { "tag": "v2021.08", "title": "Senior Associate, Data Analytics Engineering", "org": "Publicis Groupe", "unit": "Digitas Mumbai", "start": "2021-08", "end": "2022-09", "location": "Mumbai", "summary": "Cut a client's monthly cloud spend by 98% by rationalising the compute layer of the pipelines. Scaled Facebook Ads ingestion from 3 to 100+ weekly loads, taking deployment from hours to minutes. Led a team of 3 on automated pipelines and dashboards for near real-time campaign analytics.", "line": "Cut a client's monthly cloud spend by 98% by rationalising the compute layer of the pipelines. Scaled Facebook Ads ingestion from 3 to 100+ weekly loads, taking deployment from hours to minutes. Led a team of 3 on automated pipelines and dashboards for near real-time campaign analytics." }, { "tag": "v2021.06", "title": "Manager, Data and Technology", "org": "WPP", "unit": "GroupM · Mindshare", "start": "2021-06", "end": "2021-10", "location": "Mumbai", "summary": "Built a GCP data lake for Disney+ Hotstar that dropped reporting cycles from days to hours. Unified app, ad, and CRM data so content decisions could be made in near real-time.", "line": "Built a GCP data lake for Disney+ Hotstar that dropped reporting cycles from days to hours. Unified app, ad, and CRM data so content decisions could be made in near real-time." }, { "tag": "v2020.12", "title": "Senior Associate, Data Science", "org": "Publicis Groupe", "unit": "Digitas Mumbai", "start": "2020-12", "end": "2021-06", "location": "Mumbai", "summary": "Built and deployed multilingual sentiment analysis pipelines, doubling labelled-data coverage for planners. Created Power BI dashboards that spread across teams and cut manual update time by half.", "line": "Built and deployed multilingual sentiment analysis pipelines, doubling labelled-data coverage for planners. Created Power BI dashboards that spread across teams and cut manual update time by half." }, { "tag": "v2019.07", "title": "Associate, Data Science", "org": "Publicis Groupe", "unit": "Digitas Mumbai", "start": "2019-07", "end": "2020-12", "location": "Mumbai", "summary": "Automated data collection and transformation workflows, turning months of manual effort into minutes.", "line": "Automated data collection and transformation workflows, turning months of manual effort into minutes." } ], "community": [ { "name": "MERL Visitor Insight Engine", "venue": "Museum of English Rural Life · Volunteer", "year": "2025 - Present", "link": null }, { "name": "Data Science Festival Mentorship", "venue": "Data Science Festival · Mentor", "year": "2026 - Present", "link": "https://datasciencefestival.com" }, { "name": "Google Developer Specialist", "venue": "Google · Invite-only Programme", "year": "2026", "link": null } ], "education": [ { "degree": "Master of Science in Data Analytics", "org": "University of Strathclyde (Business School)", "location": "Glasgow, UK", "years": "2022 - 2023", "grade": "First-Class Honours / Distinction" }, { "degree": "Bachelor of Computer Engineering", "org": "University of Mumbai (Shah & Anchor Kutchhi Engineering College)", "location": "Mumbai, India", "years": null, "grade": "3.25 GPA" }, { "degree": "Three-year Diploma in Computer Engineering", "org": "Shri Bhagubhai Mafatlal Polytechnic (SBMP)", "location": "Mumbai, India", "years": null, "grade": "7.77 / 10" } ], "ecosystem": { "note": "Owners are named once, in the semantic layer, and access approvals, quality routing and Orca's context all reuse them.", "nodes": [ { "id": "sources", "name": "Source systems", "line": "operational systems", "kind": "people", "project": null, "state": "built" }, { "id": "raw", "name": "Raw layer", "line": "landed by pipelines", "kind": "system", "project": null, "state": "built" }, { "id": "core", "name": "Data Marts Core", "line": "dbt on Dagster", "kind": "system", "project": "Data Marts Core", "state": "built" }, { "id": "marts", "name": "Data marts", "line": "tables in BigQuery", "kind": "system", "project": null, "state": "built" }, { "id": "semantics", "name": "Orca Semantics", "line": "definitions and owners", "kind": "system", "project": "Orca Semantics", "state": "in_build" }, { "id": "orca", "name": "Orca", "line": "the AI assistant", "kind": "system", "project": "Orca", "state": "built" }, { "id": "company", "name": "The company", "line": "everyone at South Pole", "kind": "people", "project": null, "state": "built" }, { "id": "visuals", "name": "Orca Visuals", "line": "dashboards as code", "kind": "system", "project": "Orca Visuals", "state": "in_build" }, { "id": "plan", "name": "Data Foundation Plan", "line": "the strategy the executive committee adopted", "kind": "system", "project": "Data Foundation Plan", "state": "built" }, { "id": "platform", "name": "Enterprise data platform", "line": "one Terraform estate for the company", "kind": "band", "project": "Enterprise Data Platform Migration", "state": "built" }, { "id": "access", "name": "Data Access Governance", "line": "grants with a reason and an audit trail", "kind": "band", "project": "Data Access Governance", "state": "built" }, { "id": "cost", "name": "Cost Deep Dive", "line": "spend and anomalies on every project", "kind": "band", "project": "Cost Deep Dive", "state": "built" }, { "id": "quality", "name": "Data quality at scale", "line": "tests and alerts routed to owners", "kind": "band", "project": null, "state": "planned" }, { "id": "sla", "name": "Service levels", "line": "freshness and reliability targets", "kind": "band", "project": null, "state": "planned" } ], "edges": [ { "from": "sources", "to": "raw", "label": "lands in" }, { "from": "raw", "to": "core", "label": "transformed by" }, { "from": "core", "to": "marts", "label": "builds" }, { "from": "marts", "to": "semantics", "label": "defined in" }, { "from": "semantics", "to": "orca", "label": "gives meaning to" }, { "from": "marts", "to": "orca", "label": "queried as the signed-in person" }, { "from": "semantics", "to": "visuals", "label": "certified metrics for" }, { "from": "orca", "to": "company", "label": "answers" }, { "from": "visuals", "to": "company", "label": "dashboards for" }, { "from": "plan", "to": "core", "label": "sets the roadmap for everything below" } ] }, "roadmap": null, "diagrams": { "platform": { "dark": "https://saahil.co.uk/diagrams/platform-dark.svg", "light": "https://saahil.co.uk/diagrams/platform-light.svg", "title": "How it fits together", "lede": "The systems in production, what each one does and what flows between them." }, "roadmap": null }, "files": { "resume_pdf": "https://saahil.co.uk/Saahil-Mehta-Resume.pdf", "resume_json": "https://saahil.co.uk/resume.json", "cv_md": "https://saahil.co.uk/cv.md", "llms_txt": "https://saahil.co.uk/llms.txt", "status_json": "https://saahil.co.uk/status.json" } }