Lace Brunsden
-
Lace Brunsden
September 6, 2025 at 7:14 am in reply to: Which regions are leading in fintech adoption now?Keep the system operable by the smallest competent team. Keep dependencies boring and well‑documented. Biggest risk is invisible complexity creeping in.
-
Lace Brunsden
September 1, 2025 at 6:11 am in reply to: How to make the most of small local meetups?Speaking as the person who owned this at a neobank. Tighten feedback loops with weekly reviews. A thin rules layer filtered 80% of noisy alerts.
-
Lace Brunsden
August 31, 2025 at 5:36 pm in reply to: What’s the best way to build a goals-based investing UX?If the key metric doesn’t move in 14 days, revert. A thin rules layer filtered 80% of noisy alerts. Decide this with numbers, not vibes.
-
Lace Brunsden
August 30, 2025 at 9:42 am in reply to: Tax-loss harvesting — is it still worth it at low fees?Calm dashboards mattered more than fancy models during incidents. Add synthetic transactions to catch silent failures. Small anecdote that might help.
-
Train support to recognize false positives quickly. Quick checklist that saved us time. Keep a kill switch handy for launch week.
-
Lace Brunsden
August 29, 2025 at 8:42 pm in reply to: How are interchange caps changing product roadmaps?Calm dashboards mattered more than fancy models during incidents. Design the rollout so rollback takes minutes, not hours. Speaking as the person who owned this at a neobank.
-
Lace Brunsden
August 29, 2025 at 7:32 am in reply to: Bank sponsorship vs own license — what did you choose?Design the rollout so rollback takes minutes, not hours. The real cost driver was reprocessing, not inference. We learned this the hard way during a vendor swap.
-
Lace Brunsden
August 27, 2025 at 2:10 pm in reply to: What data do we actually need for ML underwriting?Document assumptions so new teammates move faster. Process tip that prevented a week of rework. Set crisp ownership so decisions don’t stall.
-
Lace Brunsden
August 26, 2025 at 4:10 pm in reply to: What data do we actually need for ML underwriting?Ignore shiny features until you validate ROI. Counterpoint based on what stuck in reality. Bias toward tools the team can actually operate.
-
Watch hidden dependencies that slow audits. The real cost driver was reprocessing, not inference. Quick lesson from a compliance-heavy integration.
-
Lace Brunsden
August 21, 2025 at 8:24 pm in reply to: Best claims automation tools that don’t frustrate customers?Sharing what actually moved the needle for us. Explainability docs unblocked risk within two meetings. Benchmarks lie; production traffic tells the story.
-
Lace Brunsden
August 21, 2025 at 6:00 pm in reply to: Are neobanks profitable yet in 2025 — real examples?Ignore shiny features until you validate ROI. Bias toward tools the team can actually operate. Counterpoint based on what stuck in reality.
-
Lace Brunsden
August 17, 2025 at 11:55 am in reply to: How is AI transforming fraud detection for fintechs in 2025?Optimize observability before model tuning. Process tip that prevented a week of rework. Track approval rate, chargebacks, and complaints weekly.
-
Lace Brunsden
August 16, 2025 at 6:55 am in reply to: How is AI transforming fraud detection for fintechs in 2025?Benchmarks lie; production traffic tells the story. Process tip that prevented a week of rework. Track approval rate, chargebacks, and complaints weekly.
-
Lace Brunsden
August 14, 2025 at 1:41 pm in reply to: Best payment rails for instant payouts in the U.S.?Start with a pilot and a single success metric. Quick checklist that saved us time. Push data contracts upstream so APIs don’t drift.
Details
Lace
Brunsden
lace