Dylan Elder
PRODUCT MANAGER | ENTERPRISE WORKFLOWS, APIs & AI
San Francisco, CA | dylankelder@gmail.com | dylankelder.com | linkedin.com/in/dylanelder
Product manager and founder who turns recurring, judgment-heavy work into products where software handles the repetition and people keep the decision. Own the NBA and college-football pricing-API roadmaps at Swish Analytics (25+ client brands). Founded and built Agency Edge, an LLM research and CRM‑enrichment product for knowledge workers where every Salesforce write waits for staff confirmation.
EXPERIENCE
Swish Analytics | Product Manager
Dec 2022–PresentOwn the NBA and college-football (CFB) real-time pricing API roadmaps for sportsbooks and prediction markets. Joined at ~50 employees (now 250+) to build in-house trader tools; expanded to API roadmap ownership in Jan 2025. Led 30+ releases across data science, engineering, trading, visualization and customer success.
- Launched live (in-play) CFB pricing after a major client took pre-match pricing in-house: opened a test API 3+ months ahead and staged the rollout, from 1 live customer at the Sep 2025 launch to 7 by Sep 2026.
- Enabled one trader per game instead of two, doubling the in-play games Swish could offer: gathered requirements with the ~30-trader desk and shipped a redesigned CFB trading view, training sandbox and in-play view (Aug 2025).
- Chose to lift restrictions in a shared pricing endpoint instead of building a client-specific one when a client’s integration stalled; the client went live, and shared pricing later extended to 26 existing CFB markets with no change to its request or response contract.
- Replaced per-customer endpoint requests with shared, optional filters on the most-used thresholds; the filters now carry 80%+ of NBA and CFB alternate-line requests with ~90% smaller responses.
- Traced two clients’ lost NBA live player selections to a market-grouping/game-state mismatch and made the case for a data-side fix that kept market IDs and client integrations unchanged; the fix shipped within a week and the client confirmed the issue looked resolved.
Agency Edge | Founder
Aug 2023–PresentFounded and built an LLM research and CRM-enrichment product for non-technical outreach and advancement teams, with shared research and review capabilities serving three paying organizations and customer-specific matching rules kept separate. Part-time alongside Swish.
- Decomposed staff research, categorization and data entry into reusable AI steps (structured outputs, function calling) with staff-set confidence thresholds for auto-approval and explicit confirmation before any Salesforce write; 88% of proposed changes accepted as written, 12% edited or rejected.
- Enriched a photography agency’s ~40,000-contact CRM in place, writing research fields and photographer-match tags back to Salesforce and Campaign Monitor: 3,000+ missing emails recovered, 6,000+ employer records updated, and customer-observed bounce rate down from 2.6% to 0.6%.
- Converted a 160-person alumni-research pilot at Marin Academy into a paid 5,078-record engagement by acting on staff feedback: added match-confidence information, filterable interest categories and preserved school identifiers for matching. For SF University High School, surfaced candidate emails for 154 of 495 alumni without one.
Monster Roster | Founder
Jan 2015–Dec 2019Raised $500K and led a 6-person engineering, data science and design team building a consumer sports-analytics optimizer: 2,300+ paying customers and the first company selected for the Philadelphia 76ers Innovation Lab.
- Redesigned the product into a guided optimizer when customers following identical recommendations began diluting each other’s results: the system proposed a core and each customer made the final choices.
SKILLS
AI & build: LLM pipelines (structured outputs, function calling); Python/Flask, SQLAlchemy, Celery/Redis, React/TypeScript; SQL
Product: Roadmap ownership, product strategy, customer discovery, workflow design, cross-functional delivery, backward‑compatible API changes, CRM data quality, review and approval design, staged rollouts
Systems: Salesforce, Campaign Monitor, Postman, Datadog, GitHub
EDUCATION
Georgetown University | B.A., Psychology | 2019–2022 (returned to complete degree)

