Costa Rica-based data architect and ML engineer at Snowflake partner Infostrux who became a 2024 Data Superhero after roughly a year of prolific Snowflake ML and LLM blogging; one of the earliest community writers on running LLMs inside Snowflake. Index: [[snowflake-data-superheroes/index|index]].
Known for
- Running LLMs inside Snowflake before Cortex existed: deploying Hugging Face models via vectorized Python UDFs (Jun 2023), managing them with the Snowpark ML Model Registry (Sep 2023), and fine-tuning on Snowpark-optimized warehouses (Oct 2023) (Medium profile, fine-tuning post).
- Cortex AI tutorials on the Snowflake Builders Blog, led by "Snowflake Cortex LLM: Self-Querying in Snowflake with Cortex AI" (May 13, 2024, 162 claps, his most-clapped Cortex post) (article).
- A 5-part "Empowering LangGraph" series (Jul 2025) wrapping the Snowflake SQL REST API, Cortex Complete, Cortex Search and Cortex Agents as LangChain/LangGraph tools, ending in an agentic query system over structured and unstructured data (Part 4, Medium profile).
Contribution channels
- Community answers: ? (no Snowflake Community forum activity found)
- User group: ? (no user-group leader or speaker page found)
- Blog: Primary channel. About 29 posts Jan 2023 to Jul 2025 on the Infostrux Engineering Blog and the Snowflake Builders Blog (medium.com/snowflake), all under Medium handle @fahernandez (profile). Builders Blog examples: Cortex self-querying (May 2024), LLMOps case study with the Snowpark ML Model Registry (Jun 2024), Cortex Search activity planner (Oct 2024), 2-part Cortex + Streamlit sports assistant (Part 2), Snowpark + FastAPI and custom OAuth client (OAuth post), and the LangGraph series.
- YouTube / talks: ? (none found; a "LangGraph and Snowflake Cortex AI" video on Snowflake Developers is by a different presenter, not him)
- Open source: GitHub fahernandez (Costa Rica, ~48 repos per search snippet; no Snowflake-specific flagship repo identified).
- Books: ?
- Certs: Per LinkedIn: SnowPro Core (Dec 2022), SnowPro Advanced Data Scientist (Jan 2023), dbt Developer (Jun 2023), SnowPro Advanced Data Analyst (Nov 2023), SnowPro Advanced Architect (Mar 2025), plus AWS SA Associate and Databricks Spark associate; LinkedIn also lists "Data Superhero 2024 (Feb 2024)" (LinkedIn).
Path to recognition
- Dec 2022 to Jan 2023: earned SnowPro Core and SnowPro Advanced Data Scientist (LinkedIn).
- Jan 2023: opened with a burst of 7 Infostrux Engineering Blog posts in about four weeks on ML in Snowflake (forecasting with vectorized UDFs, a homemade model registry, SageMaker Autopilot and OpenAI via external functions, a data science quickstart) (Medium profile). This is his first visible Snowflake-community activity.
- Jun 5, 2023: first Snowflake Builders Blog post, "Deploying pre-trained LLMs in Snowflake" (180 claps, his most-clapped post) (Medium profile). Getting picked up by Snowflake's own publication, on the hottest topic of 2023, is plausibly the step that put him on the program's radar (inference).
- Sep 2023: two more Builders Blog LLM posts on the Snowpark ML Model Registry; Oct to Nov 2023: LLM fine-tuning and an end-to-end AI case (dbt + Hugging Face + MLOps) on the Infostrux blog (Medium profile).
- Feb 8, 2024: named one of 20 NEW Data Superheroes ("Fabian Hernandez, Costa Rica") (2024 announcement). Roughly 13 months, ~20 posts and 3 SnowPro certs after his first post.
- After selection he shifted to Cortex-native content (May 2024 onward) and agent orchestration (Jul 2025), which lines up with his renewals in 2025 and 2026 (inference).
- Pattern (inference): recognition came from written technical depth on a single emerging theme (ML/LLMs inside Snowflake), not from events or forum answers. No talks or user-group roles are visible.
Notes
- Directory card (scraped 2026-09-29): Data Architect, Infostrux. Member-since tag: 2024 (matches the Feb 2024 "new" announcement). On the 2025 Medium roster. 2026 membership presumed from the current directory; directory (130) does not exactly equal the announced 2026 cohort (125). See [[_roster-notes]].
- Authorship verified: the Cortex self-querying post's byline and author link resolve to medium.com/@fahernandez, bio "Data Architect @ Infostrux. Computer Scientist, Statistician and Machine Learning Engineer." (article).
- Identity: common name. Medium handle, GitHub handle (fahernandez, Costa Rica), the cr.linkedin.com profile, and the 2024 announcement all point to one Costa Rica-based person; other same-name LinkedIn profiles (robotics, workers' comp, Sogeti) are different people.
- Employer conflict: search snippets title him "Field CTO Data Science & ML at Infostrux Solutions" and mention prior AI research at the University of Nottingham (MSc 2021-22 per LinkedIn). A LinkedIn fetch (2026-09-29) showed a current role at WHOOP. Employer kept as Infostrux per the Snowflake directory; a move to WHOOP (end-customer) is possible but unconfirmed. If confirmed, employer_type would become end-customer.
- No Medium posts after Jul 25, 2025 are visible, consistent with a job change (inference).
- For Ben: closest analog of the partner-consultant path. Blog on one hot theme on the employer blog, get syndicated to the Builders Blog, stack SnowPro Advanced certs, apply (2024 announcement).
Sources
- Snowflake Data Superheroes directory, Americas
- Celebrating the 2025 Data Superheroes (Medium roster)
- Introducing the 2024 Snowflake Data Superheroes (Feb 8, 2024)
- Medium author profile @fahernandez
- Snowflake Cortex LLM: Self-Querying (May 2024)
- Empowering LangGraph Part 4: Cortex Agents (Jul 2025)
- Cortex AI + Streamlit sports assistant Part 2 (Nov 2024)
- Custom OAuth client with FastAPI (Jul 2025)
- Fine-tuning LLMs on Snowpark-optimized warehouses (Infostrux blog, Oct 2023)
- LinkedIn profile
- GitHub fahernandez