open to senior & staff roles
milpitas, ca · remote-friendly

Analytics Engineering Leader /
building fintech
data platforms.

15+ years at Workday, Lyft, and beyond — now building what fintech lending teams actually need: end-to-end data models, SPV reconciliation, risk analytics, and investor-grade reporting. My flagship project LoanLens models 10,000 loans and 1.35M payment events with <0.1% reconciliation tolerance.

1.35M
payment events modeled (LoanLens)
<0.1%
reconciliation tolerance achieved
3
SPV covenant monitoring systems
15+
years in enterprise analytics
Previously at
why fintech

Built for reconciliation,
risk, and investor trust.

Fintech data teams face a specific set of problems: loan portfolios that must reconcile to the cent, covenant monitoring that cannot miss, and investors who need to trust the numbers before they fund the next facility. I've spent 15 years building data systems that earn that trust.

At Workday I lead BI Analytics for a $8B+ marketing org — the discipline of modeling revenue attribution, attribution windows, and budget pacing maps directly to lending portfolio analytics. At Lyft I built growth data infrastructure at scale. Before that, supply chain analytics at Intuitive Surgical, where every number had regulatory implications.

LoanLens, my most recent build, is a simulated Series C lender data platform: 10,000 loans, 1.35M payment events, 3 SPVs modeled in dbt, reconciled end-to-end, with Claude-narrated investor memos. That's the kind of data infrastructure fintech teams need, and I can build it.

Shrikant Lambe
selected work · fintech focus

Lending analytics from loan tape to investor memo.

All 8 projects →
08 / loanlens
FinTech · Data Platform

LoanLens — portfolio intelligence for a simulated Series C lender.

10,000 loans, 1.35M payment events, 3 SPVs. Modeled in dbt, reconciled end-to-end to <0.1% tolerance, narrated by Claude. SPV covenant monitoring, 36-cohort vintage analysis, anomaly detection, and investor-grade PDF memo export. Designed and shipped as a two-day sprint to prove what a senior engineer with the right stack can build fast.

📊 1.35M payment events · 3 SPVs reconciled to <0.1% · investor-grade PDF memo in one click
1.35M
Payment events
<0.1%
Recon tolerance
36
Cohort vintages
Claude SonnetDuckDBdbt Core SnowflakeStreamlitPython
loanlens · spv-b portfolio snapshot
SPV-B · Q4 2025 · AS-OF 2026-04-22
$284M
AUM
+3.1% QoQ
3.2%
Default rate
+40bps
0.04%
Recon delta
within tol.
🤖 investor memo · claude
SPV-B at 87% facility utilization — recommend 60-day drawdown pause. 2023 vintage cohorts outperforming 12-month default curve by 40bps.
07 / sentinel
Multi-Agent · DataOps

Pipeline Sentinel — a self-healing agent for Airflow.

In fintech, a broken pipeline means delayed reconciliation, stale investor reports, and manual remediation at 2am. Pipeline Sentinel eliminates that: five specialized agents diagnose, blast-radius-scope, and remediate Airflow failures in ~11s. Pattern memory prevents repeat incidents. Confidence gate ensures only low-risk fixes auto-apply; everything else escalates with a full audit trail.

⚡ Resolves pipeline failures in ~11s · 94% median confidence · zero human pages on low-risk incidents
5
Agents in loop
94%
Median confidence
11s
Resolution time
Claude SonnetLangGraphLangSmith FastAPIAirflowStreamlit
pipeline-sentinel · agent dashboard
⚠ INCIDENT · loan_recon_pipeline · reconcile_spv_b
Monitor · task failure detectedT+04s
Diagnosis · upstream schema drift on `payment_status`T+07s
Blast Radius · 2 downstream SPV tasks identifiedT+08s
Remediation · applying `reload_schema` strategyT+09s
🛡 self-healed · confidence 94%
Schema reloaded, reconciliation tasks re-queued. No human intervention. Recon report back on schedule.
05 / cortex
Snowflake · Enterprise AI

AI-Native Data Platform on Snowflake Cortex.

A production-grade architecture blueprint for running governed AI natively inside Snowflake — zero data egress, end-to-end policy enforcement. Relevant for fintech teams that need to query loan performance and risk signals in natural language while keeping data in-warehouse for compliance. Cortex LLM functions, Cortex Search, Cortex Analyst with live interactive demos.

🏗 Zero data egress · end-to-end governed AI · live interactive demo on GitHub Pages
3
Cortex layers
0
Data egress
E2E
Governed AI
Snowflake CortexCortex Search Cortex AnalystAnthropic APISQL
snowflake-cortex-architecture
IN-WAREHOUSE GOVERNED AI · FINTECH USE CASE
Which loan cohorts have default rates above 3.5% in the last 90 days?
Q3-2024 and Q4-2024 originations showing 3.8% and 4.1% default rates respectively. SPV-A has 12 loans breaching covenant thresholds — recommend covenant review before next investor call.
01 / churn
ML · Predictive Analytics

Customer Churn — ML prediction with AI decision layer.

Traditional churn models stop at a probability score. This one goes further: SHAP-explainable predictions layered with an OpenAI-powered decision copilot that translates model outputs into retention actions. In fintech terms: the same pattern applied to delinquency prediction, early payoff risk, or portfolio segment scoring.

🎯 SHAP-explained predictions + AI decision layer · production API with FastAPI + GitHub Actions CI
SHAP
Explainability
FastAPI
Production API
CI/CD
GitHub Actions
scikit-learnOpenAISHAP StreamlitFastAPI
churn-prediction · decision copilot
RISK SCORING · BORROWER SEGMENT ANALYSIS
87%
Churn probability
high risk
0.83
AUC-ROC
strong signal
3
Top SHAP features
explainable
🤖 ai recommendation
Top signal: missed payment + 60d inactivity + declining balance. Recommend proactive outreach before next billing cycle. Similar segment — 34% retained with early intervention.
◆ let's talk fintech data

Let's build data you can lend on.

I'm actively talking to fintech teams building the next generation of lending infrastructure, portfolio analytics, and investor-grade data platforms. I can move fast and I know how the numbers need to close.

analytics engineering lead data platform architect AI engineering manager