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

Analytics Engineering Leader /
building agentic
AI systems.

15+ years in enterprise data, now focused on what's above the dashboard: multi-agent systems built on Claude, autonomous DataOps that self-heals, and AI decision layers grounded in governed data. I build on the Claude API in production — not as a demo, but as the reasoning engine for systems that run without human intervention.

3
production Claude-based agentic systems
5
specialized agents in Pipeline Sentinel
8
AI/ML systems shipped across all projects
15+
years in enterprise analytics
Previously at
building on claude

Governed data +
agentic AI systems.

I'm a practitioner, not a prompt engineer. I use the Claude API as the reasoning backbone for multi-agent systems that coordinate through shared state, respect confidence thresholds, and escalate gracefully when they're uncertain. I write about what it actually takes to build safely and reliably on top of LLMs.

At Workday I lead BI Analytics and shipped an AI Companion — natural-language queries over marketing performance data, grounded in a governed Snowflake semantic layer. The same architecture principles apply to any system where you need LLM reasoning to respect data contracts and organizational governance.

My three Claude-native projects — Pipeline Sentinel (autonomous DataOps), Growth Intelligence Agent (agentic analytics), and LoanLens (AI-narrated investor memos) — are production-grade systems with real confidence gates, audit trails, and fallback logic. That's the kind of engineering judgment I bring to agentic systems work.

Shrikant Lambe
selected work · agentic AI focus

Claude-native systems. Production-grade.

All 8 projects →
07 / sentinel
Multi-Agent · DataOps

Pipeline Sentinel — a self-healing agent for Airflow.

Five specialized agents — Monitor, Diagnosis, Blast Radius, Remediation, Reflection — coordinate through shared state to diagnose and resolve Airflow failures without paging a human. Pattern memory prevents repeat incidents. A confidence gate auto-remediates only low-risk, high-signal cases; everything else escalates with a full audit trail. Built on Claude Sonnet + LangGraph ReAct loop.

⚡ Self-heals pipeline failures in ~11s · 94% median confidence · zero human pages on low-risk incidents
5
Agents in ReAct loop
94%
Median confidence
11s
Time-to-resolution
Claude SonnetLangGraphLangSmith FastAPIAirflowStreamlit
pipeline-sentinel · agent dashboard
⚠ INCIDENT · customer_churn_pipeline · transform_features
Monitor · task failure detectedT+04s
Diagnosis · upstream schema drift on `plan_tier`T+07s
Blast Radius · 3 downstream tasks identifiedT+08s
Remediation · applying `reload_schema` strategyT+09s
🛡 self-healed · confidence 94%
Schema reloaded, 3 tasks re-queued. No human intervention. Similar incident 2026-04-03 — same fix, resolved in 11s.
06 / growth-agent
Agentic AI · SaaS Analytics

Growth Intelligence Agent — a virtual revenue analyst.

Autonomously monitors SaaS growth metrics across 6 categories, detects anomalies with severity scoring, and surfaces RAG-grounded strategic recommendations from company playbooks through a natural-language interface. Built on Claude 3.5 Sonnet + LangChain. Demonstrates the pattern of an LLM-as-analyst that cites sources, scores its own confidence, and stays grounded in real data.

🤖 Autonomous anomaly detection + playbook-grounded recommendations · natural-language interface
6
Metric categories
FAISS
Playbook retrieval
NL
Query interface
Claude 3.5 SonnetLangChainFAISS RAGStreamlitPlotly
growth-intelligence-agent · week 17
SAAS PORTFOLIO · WEEK 17 · REAL-TIME
$2.4M
MRR
+6.2% WoW
118%
NRR
above target
↑2.1%
Churn risk
anomaly
ANOMALY · HIGH — enterprise churn spiked +2.1pp WoW · 3 accounts flagged
Why did enterprise churn jump this week?
Per ICP playbook §3.2: accounts with <60% feature adoption + no QBR in 90d = highest signal. 3 flagged accounts match both. Recommend CSM outreach this week.
08 / loanlens
FinTech · AI-Narrated Analytics

LoanLens — Claude-narrated investor intelligence.

A fintech data platform where Claude serves as the analyst: it reads the portfolio state, identifies covenant risks, and generates investor-grade memo narratives with structured recommendations. 1.35M payment events, 3 SPVs modeled in dbt + DuckDB. Demonstrates how Claude can be grounded in structured data outputs to produce trustworthy, citation-backed analysis.

📊 Claude-narrated investor memos · 1.35M payment events · 3 SPVs reconciled to <0.1%
1.35M
Payment events
Claude
Narrative engine
36
Cohort vintages
Claude SonnetDuckDBdbt Core SnowflakeStreamlitPython
loanlens · investor memo · claude
CLAUDE-GENERATED INVESTOR NARRATIVE · SPV-B
🤖 claude · portfolio analysis · Q4 2025
SPV-B at 87% facility utilization — recommend 60-day drawdown pause. 2023 vintage cohorts outperforming 12-month default curve by 40bps. Reconciliation delta at 0.04%: within tolerance. Q1 2026 covenant check passed on all 3 metrics.
$284M
AUM
+3.1% QoQ
3.2%
Default rate
+40bps
0.04%
Recon delta
within tol.
05 / cortex
Snowflake · Enterprise AI

AI-Native Data Platform on Snowflake Cortex.

A production-grade architecture for running governed AI natively inside Snowflake, with Anthropic API integration for hybrid LLM use cases. Demonstrates the pattern of keeping sensitive data in-warehouse while using external LLMs for advanced reasoning — an important constraint in enterprise AI deployments where data residency and governance matter.

🏗 In-warehouse AI + Anthropic API integration · zero data egress · live interactive demo
3
Cortex layers
0
Data egress
Hybrid
LLM architecture
Snowflake CortexCortex Analyst Anthropic APISQL
snowflake-cortex-architecture
HYBRID LLM ARCHITECTURE · ANTHROPIC API
-- hybrid: in-warehouse + Anthropic API reasoning SELECT customer_id, SNOWFLAKE.CORTEX.COMPLETE( 'claude-3-haiku', 'Analyze this customer segment and identify churn signals: ' || segment_summary ) AS ai_insight FROM customer_segments WHERE risk_score > 0.7
◆ anthropic api · grounded reasoning
Data stays in Snowflake. Claude reasons over pre-aggregated summaries — no raw PII leaves the warehouse. Governance preserved, reasoning unlocked.
◆ let's talk agentic AI

Let's build AI systems that ship.

I'm actively talking to teams building the next generation of agentic AI infrastructure — systems that coordinate multiple models, ground reasoning in governed data, and escalate gracefully when uncertain. I build on Claude in production and I know what it takes to make these systems reliable.

analytics engineering lead AI engineering manager senior analytics engineer