Siva NekkantiPersonal website for Siva Nekkanti, a data engineer focused on ML infrastructure, privacy engineering, and reliable data systems.https://www.sivanekkanti.com/Learning From My First Broken Pipelinehttps://www.sivanekkanti.com/blog/2018-q1-learning-pipeline-failure/https://www.sivanekkanti.com/blog/2018-q1-learning-pipeline-failure/Early notes on why a data pipeline that runs is not always a data pipeline that works.Mon, 12 Feb 2018 00:00:00 GMTSchemas Are Promiseshttps://www.sivanekkanti.com/blog/2018-q2-schemas-as-promises/https://www.sivanekkanti.com/blog/2018-q2-schemas-as-promises/Why I started treating schemas as contracts between teams, not just database metadata.Wed, 09 May 2018 00:00:00 GMTDebugging With Row Countshttps://www.sivanekkanti.com/blog/2018-q3-debugging-with-row-counts/https://www.sivanekkanti.com/blog/2018-q3-debugging-with-row-counts/A simple reliability habit that helped me understand data movement before adding bigger tools.Tue, 14 Aug 2018 00:00:00 GMTBatch Jobs Need Ownershttps://www.sivanekkanti.com/blog/2018-q4-batch-jobs-and-ownership/https://www.sivanekkanti.com/blog/2018-q4-batch-jobs-and-ownership/A reflection on why scheduled jobs become fragile when nobody clearly owns their behavior.Fri, 16 Nov 2018 00:00:00 GMTData Quality Is Product Qualityhttps://www.sivanekkanti.com/blog/2019-q1-data-quality-is-product-quality/https://www.sivanekkanti.com/blog/2019-q1-data-quality-is-product-quality/How I began connecting backend data issues to the product experiences people actually see.Mon, 11 Feb 2019 00:00:00 GMTBackfills Taught Me Humilityhttps://www.sivanekkanti.com/blog/2019-q2-backfills-taught-me-humility/https://www.sivanekkanti.com/blog/2019-q2-backfills-taught-me-humility/What historical data corrections revealed about assumptions hidden inside pipelines.Mon, 13 May 2019 00:00:00 GMTPartitions and Practical Performancehttps://www.sivanekkanti.com/blog/2019-q3-partitions-and-practical-performance/https://www.sivanekkanti.com/blog/2019-q3-partitions-and-practical-performance/A note on learning to make warehouse performance understandable instead of mysterious.Mon, 12 Aug 2019 00:00:00 GMTDocumenting the Weird Partshttps://www.sivanekkanti.com/blog/2019-q4-documenting-the-weird-parts/https://www.sivanekkanti.com/blog/2019-q4-documenting-the-weird-parts/Why the most useful data documentation often explains exceptions, not happy paths.Mon, 18 Nov 2019 00:00:00 GMTReliability During Uncertaintyhttps://www.sivanekkanti.com/blog/2020-q1-reliability-during-uncertainty/https://www.sivanekkanti.com/blog/2020-q1-reliability-during-uncertainty/How changing business conditions made me think harder about freshness, drift, and operational signals.Mon, 17 Feb 2020 00:00:00 GMTIdempotency as Calmhttps://www.sivanekkanti.com/blog/2020-q2-idempotency-as-calm/https://www.sivanekkanti.com/blog/2020-q2-idempotency-as-calm/Why rerunnable jobs made data operations feel less dramatic.Mon, 11 May 2020 00:00:00 GMTData Contracts Before They Were Fashionablehttps://www.sivanekkanti.com/blog/2020-q3-data-contracts-before-they-were-fashionable/https://www.sivanekkanti.com/blog/2020-q3-data-contracts-before-they-were-fashionable/How producer-consumer expectations became a recurring theme in my data engineering work.Mon, 10 Aug 2020 00:00:00 GMTObservability for Humanshttps://www.sivanekkanti.com/blog/2020-q4-observability-for-humans/https://www.sivanekkanti.com/blog/2020-q4-observability-for-humans/A reflection on alerts, dashboards, and making data systems easier to operate.Mon, 09 Nov 2020 00:00:00 GMTFeature Pipelines Are Data Productshttps://www.sivanekkanti.com/blog/2021-q1-feature-pipelines-are-data-products/https://www.sivanekkanti.com/blog/2021-q1-feature-pipelines-are-data-products/How ML feature work changed the way I thought about ownership and interfaces.Mon, 08 Feb 2021 00:00:00 GMTPrivacy Is a Systems Propertyhttps://www.sivanekkanti.com/blog/2021-q2-privacy-is-a-systems-property/https://www.sivanekkanti.com/blog/2021-q2-privacy-is-a-systems-property/Why privacy engineering belongs in architecture decisions, not only review checklists.Mon, 10 May 2021 00:00:00 GMTLineage as a Debugging Toolhttps://www.sivanekkanti.com/blog/2021-q3-lineage-as-a-debugging-tool/https://www.sivanekkanti.com/blog/2021-q3-lineage-as-a-debugging-tool/How lineage became more useful to me when I stopped treating it as a catalog feature.Mon, 09 Aug 2021 00:00:00 GMTTesting Transformations Without Pretending Data Is Codehttps://www.sivanekkanti.com/blog/2021-q4-testing-transformations/https://www.sivanekkanti.com/blog/2021-q4-testing-transformations/A more practical view of testing analytics and pipeline logic.Mon, 08 Nov 2021 00:00:00 GMTML Infrastructure Is a Feedback Systemhttps://www.sivanekkanti.com/blog/2022-q1-ml-infrastructure-is-a-feedback-system/https://www.sivanekkanti.com/blog/2022-q1-ml-infrastructure-is-a-feedback-system/Why I started thinking beyond training pipelines and toward learning loops.Mon, 07 Feb 2022 00:00:00 GMTServing Features Reliablyhttps://www.sivanekkanti.com/blog/2022-q2-serving-features-reliably/https://www.sivanekkanti.com/blog/2022-q2-serving-features-reliably/A note on the gap between offline feature logic and production serving expectations.Mon, 09 May 2022 00:00:00 GMTDeleting Data Is Engineeringhttps://www.sivanekkanti.com/blog/2022-q3-deleting-data-is-engineering/https://www.sivanekkanti.com/blog/2022-q3-deleting-data-is-engineering/Why retention, deletion, and lifecycle controls deserve real design attention.Mon, 08 Aug 2022 00:00:00 GMTCost Is an Observability Signalhttps://www.sivanekkanti.com/blog/2022-q4-cost-is-an-observability-signal/https://www.sivanekkanti.com/blog/2022-q4-cost-is-an-observability-signal/How cloud spend helped me see inefficient data systems before they became incidents.Mon, 07 Nov 2022 00:00:00 GMTData Platforms Need Product Thinkinghttps://www.sivanekkanti.com/blog/2023-q1-data-platforms-need-product-thinking/https://www.sivanekkanti.com/blog/2023-q1-data-platforms-need-product-thinking/Why internal platforms should be designed around workflows, not just capabilities.Mon, 06 Feb 2023 00:00:00 GMTRetrieval Is Data Engineeringhttps://www.sivanekkanti.com/blog/2023-q2-retrieval-is-data-engineering/https://www.sivanekkanti.com/blog/2023-q2-retrieval-is-data-engineering/How LLM applications made familiar data quality problems show up in a new interface.Mon, 08 May 2023 00:00:00 GMTEvaluating LLM Productshttps://www.sivanekkanti.com/blog/2023-q3-evaluating-llm-products/https://www.sivanekkanti.com/blog/2023-q3-evaluating-llm-products/A practical view of evaluation data, feedback loops, and product quality.Mon, 07 Aug 2023 00:00:00 GMTPrivacy in AI Systemshttps://www.sivanekkanti.com/blog/2023-q4-privacy-in-ai-systems/https://www.sivanekkanti.com/blog/2023-q4-privacy-in-ai-systems/Why AI products make data minimization and permission boundaries even more important.Mon, 06 Nov 2023 00:00:00 GMTReliable Data Systems Have Memoryhttps://www.sivanekkanti.com/blog/2024-q1-reliable-data-systems-have-memory/https://www.sivanekkanti.com/blog/2024-q1-reliable-data-systems-have-memory/A more mature view of run history, decisions, and institutional knowledge.Mon, 05 Feb 2024 00:00:00 GMTDesigning for Data Minimizationhttps://www.sivanekkanti.com/blog/2024-q2-designing-for-data-minimization/https://www.sivanekkanti.com/blog/2024-q2-designing-for-data-minimization/How privacy constraints can lead to cleaner, more purposeful data architecture.Mon, 06 May 2024 00:00:00 GMTAI Systems Need Data Contracts Toohttps://www.sivanekkanti.com/blog/2024-q3-ai-systems-need-data-contracts-too/https://www.sivanekkanti.com/blog/2024-q3-ai-systems-need-data-contracts-too/Why prompts, retrieval context, and evaluation data need explicit interfaces.Mon, 05 Aug 2024 00:00:00 GMTFrom Pipelines to Platformshttps://www.sivanekkanti.com/blog/2024-q4-from-pipelines-to-platforms/https://www.sivanekkanti.com/blog/2024-q4-from-pipelines-to-platforms/How my perspective shifted from building individual workflows to enabling teams.Mon, 04 Nov 2024 00:00:00 GMTMeasuring Retrieval Qualityhttps://www.sivanekkanti.com/blog/2025-q1-measuring-retrieval-quality/https://www.sivanekkanti.com/blog/2025-q1-measuring-retrieval-quality/A deeper look at evaluating the data layer behind LLM product behavior.Mon, 03 Feb 2025 00:00:00 GMTThe Hard Part of Reliable AIhttps://www.sivanekkanti.com/blog/2025-q2-the-hard-part-of-reliable-ai/https://www.sivanekkanti.com/blog/2025-q2-the-hard-part-of-reliable-ai/Why reliable AI products depend on ordinary engineering discipline around changing systems.Mon, 05 May 2025 00:00:00 GMTBuilding Trustworthy Feature Storeshttps://www.sivanekkanti.com/blog/2025-q3-building-trustworthy-feature-stores/https://www.sivanekkanti.com/blog/2025-q3-building-trustworthy-feature-stores/What feature infrastructure needs beyond a place to store feature values.Mon, 04 Aug 2025 00:00:00 GMTData Engineering for Agentic Productshttps://www.sivanekkanti.com/blog/2025-q4-data-engineering-for-agentic-products/https://www.sivanekkanti.com/blog/2025-q4-data-engineering-for-agentic-products/How tool-using AI products raise the bar for logs, permissions, and evaluation data.Mon, 03 Nov 2025 00:00:00 GMTLessons From Reliable Data Systemshttps://www.sivanekkanti.com/blog/2026-q1-lessons-from-reliable-data-systems/https://www.sivanekkanti.com/blog/2026-q1-lessons-from-reliable-data-systems/A synthesis of what years of pipeline, privacy, and ML infrastructure work have taught me.Mon, 02 Feb 2026 00:00:00 GMTWhere Data Engineering Meets AI Product Engineeringhttps://www.sivanekkanti.com/blog/2026-q2-where-data-engineering-meets-ai-product-engineering/https://www.sivanekkanti.com/blog/2026-q2-where-data-engineering-meets-ai-product-engineering/My current view on the overlap between reliable data platforms and useful AI products.Mon, 04 May 2026 00:00:00 GMT