Srujan Gowda

Program Manager with ~3 years of experience driving AI automation, data platform development, and internal product ownership. Designed and shipped a production-grade AI data engine at Tracxn - covering end-to-end extraction, LLM-based classification, enrichment, and validation across 30,000+ global company websites. Known for translating ambiguous requirements into scalable systems, owning roadmaps end-to-end, and working across engineering and business teams to ship.

Program Manager – Internal Product at Tracxn Technologies (2024-11 – Present)

• AI Data Engine - Designed and owned the full product lifecycle of an end-to-end automation engine processing 30,000+ global company websites - covering domain validation, canonical redirect resolution, homepage scraping, LLM-based URL categorisation, structured JSON extraction, and multi-stage enrichment against internal APIs, with confidence scoring and evidence logging at each stage.
• LLM Pipeline & Prompt Engineering - Built a two-pass LLM system where the first pass categorises URLs (Basic, Contact, People, Portfolio, Careers) and the second extracts structured company-level data from scraped markdown into validated JSON. Iteratively refined prompts to reduce hallucinations and improve extraction accuracy.
• Enrichment & Validation Layer - Developed fuzzy matching logic to cross-reference extracted records (company names, CXO profiles, locations, portfolios) against existing platform datasets - with confidence scores on a 0–1 scale and evidence quotes stored per field for full traceability.
• CXO Profiling System - Spearheaded a CXO profile updater using context injection and classification logic to generate structured JSON outputs for executive profiles across 30,000+ companies, reducing manual research effort by 92%.
• GCP Spend Alert Bot - Built and deployed a Google Chat bot on Cloud Run that pulls daily GCP billing data, computes 14-day rolling spend baselines, detects anomalies, and sends automated alerts - eliminating manual cost monitoring and catching overruns before month-end.
• Pipeline Review Console & MIS - Deployed an internal Cloud Run dashboard for QA sign-off on extraction runs and enrichment validation; built real-time dashboards tracking scraping success rates, LLM token costs, and API call metrics - contributing to 90%+ reduction in manual effort and giving leadership clear program visibility.

Business Analyst at Heraizen Technologies Pvt. Ltd (2023-02 – 2024-10)

• Medical University Integration - Led end-to-end delivery for Heraizen's first medical university client - a greenfield integration across Examination and Finance modules on a config-based ERP platform previously built only for engineering colleges. Mapped domain-specific requirements from scratch and coordinated across a 6-person team to deliver on time.
• Requirement Engineering - Acted as single point of contact between university stakeholders and the dev team: gathered requirements across academics, finance, and examination workflows, documented functional specs, and raised change requests for gaps in the existing config system.
• QA & Client Handoff - Owned testing and validation for each module before client handoff - reproducing edge cases flagged by university admins, coordinating fixes with the dev team, and ensuring the delivered system matched agreed requirements.

B.E. in Civil Engineering – Sahyadri College of Engineering and Management (2018 – 2022)