Sri Panchavati

Enterprise Career

Where I Lead

24+ years directing enterprise architecture, cloud modernization, and AI/ML operationalization at Canada's top financial institutions. Big 4 consulting pedigree meets hands-on technical execution.

Director, Enterprise Architecture & Data Analytics

CIBC Private Wealth

May 2022 – Present · Toronto

Directed end-to-end architectural redesign and migration of 12 critical on-prem wealth apps to Azure/Databricks cloud ecosystem
Formed and chaired divisional AI committee — architecting governed adoption of Agentic AI + Microsoft Co-Pilot across enterprise workflows
Scaled organization from 4 to 50+ across multiple delivery teams — stood up offshore capabilities and built a hybrid onshore/offshore model from the ground up
Primary architectural liaison for C-suite executives — translating multi-million-dollar technology investments into core business strategy
Deployed Power BI as enterprise standard, integrating W360 portfolio management data into C-level executive dashboards
Owned budget allocation, vendor negotiations, and continuous alignment with executive stakeholders

Leader, Data Analytics & Automation Practices

Deloitte Canada

Jan 2015 – Apr 2022 · Toronto

Program Architect, Scotiabank — Orchestrated 15+ concurrent automation and data projects in the Automation Factory
Practice Lead, Workday Prism — Led Deloitte's Prism Analytics practice — defined architecture, roadmaps, training curricula, and go-to-market
Delivery Lead, Predictive Systems — Led cross-functional data science + engineering teams building enterprise-grade workforce forecasting models
Served as C-suite advisory liaison across Scotiabank, managing steering committees, program finances, and enterprise-level risk mitigation

Associate IT Architect

IBM

2010 – 2014 · India & Canada

Designed BI reporting architectures for TD Bank, HSBC, and European financial clients
Led Cognos migration and data warehouse optimization initiatives
Full-stack BI development across Cognos, Hyperion, DB2, and MS SQL Server

Senior Engineer

Accenture / Polaris · Genpact / NIIT

2003 – 2010 · India

Delivered enterprise reporting, ETL frameworks, and data migration solutions for Toyota, Wartsila, and GE
DataStage ETL development, SAP HANA, Oracle data warehouse design
Progressive technical leadership across global consulting firms

By day, I architect enterprise cloud transformations for one of Canada's largest banks. By night, I ship production AI systems.

Azure AI Fundamentals Azure Data Fundamentals Azure Fundamentals Workday Prism Analytics Workday Composite Reporting Workday Advanced Reporting

Expertise

What I Bring to the Table

The intersection of enterprise architecture discipline and hands-on AI product development. I don't just design systems on whiteboards — I build and operate them.

Data & Analytics

Enterprise Data Strategy

Lakehouse architecture (Azure/Databricks), dimensional data modelling, ETL/ELT pipeline design, enterprise BI (Power BI, Workday), data governance (PIPEDA compliance), semantic search with pgvector & vector embeddings

Architecture

Cloud & System Design

Multi-service orchestration (15+ microservices), event-driven architecture, 4-layer memory systems, graph databases (Neo4j), caching strategies (Redis ring buffers), CI/CD pipelines (Azure DevOps), infrastructure-as-code

AI / ML

Operationalized AI

Multi-agent orchestration, LLM routing & fallback chains, RAG pipelines, embedding-based retrieval, Ebbinghaus decay scoring, MCP protocol, prompt engineering, AI governance & ethics frameworks

Leadership

Executive & Organizational

Scaled teams 4→37, P&L ownership (multi-million-dollar portfolios), global delivery models (onshore/offshore), C-suite advisory, board presentations, practice generation from zero, vendor & contract management

Data Modelling

Relational & Graph

Star/snowflake schemas, temporal bi-directional graphs, multi-tenant SaaS data models, vector embedding schemas (1024-dim), tenant-isolated security models, PostgreSQL extensions (pgvector, pg_trgm)

Consulting

Big 4 & Functional

7+ years at Deloitte serving Canada's largest banks, plus IBM, Accenture, and Genpact. Pre-sales, technical proofs-of-concept, practice leadership (Workday Prism), rapid turnaround in chaotic environments.

Industries

Where I've Operated

Deep domain expertise across financial services and technology consulting, with product ventures in AI infrastructure, education, and digital media.

🏦

Banking & Wealth

CIBC Private Wealth, Scotiabank — core banking data, portfolio management, regulatory compliance

📊

Management Consulting

Deloitte — practice generation, enterprise transformation, pre-sales engineering

🤖

AI Infrastructure

Smara (Memory API), MCP Doctor, Lumen — developer tools, agent memory, protocol compliance

📚

EdTech & Digital Media

PJ Tales (AI storytelling), InBlue (financial education) — content pipelines, TTS, curriculum design

Products

What I've Built

Production-hardened systems running on real infrastructure, serving real users. Not demos. Not tutorials. Shipped products.

Lumen live

Family AI operating system with 10 specialized agents, 4-layer memory architecture (Redis, pgvector, Neo4j, Ebbinghaus decay), and multi-channel communication via WhatsApp, Telegram, and web dashboard. Local-first with Ollama, cloud fallback chain.

TypeScript Next.js PostgreSQL pgvector Neo4j Redis Ollama Cloudflare Tunnel

15+microservices

10AI agents

4-layermemory stack

4LLM providers

Smara live

Memory-as-a-Service API for AI agents. Temporal Memory Scoring with Ebbinghaus decay curves. SDKs for Python, JS/TS, MCP server, LangChain, and LlamaIndex. 96% cheaper than competitors.

Fastify PostgreSQL pgvector Voyage AI Stripe Railway MCP

--status

3-callintegration

$19dev/mo

smara.io API Docs GitHub

Scoper shipped open source

AI project estimation platform. Upload a BRD/SOW → 9 AI agents run a visual drag-and-drop pipeline → complete project estimate in under 2 minutes.

React Vite React Flow FastAPI SQLite YAML Agents

9AI agents

<2 minestimation

5workflow templates

GitHub

PJ Tales live

AI bedtime story production pipeline. Full screenplay format (Pixar/Ghibli quality), 14 TTS voices, 8-dimension quality scoring, visual pipeline IDE.

FastAPI React Flow Orpheus TTS Edge TTS Ollama FFmpeg

14TTS voices

8quality dims

MCP Doctor shipped

MCP server compliance testing and monitoring. 50+ automated checks against the official MCP spec. A-F grading system, CI/CD ready.

Fastify Next.js MCP Protocol

50+compliance checks

A-Fgrading

mcpdoctor.ai

InBlue live

Covered call ETF investment guide. 5-tab dashboard with strategies, ETF reference, wisdom library, portfolio builder, and DRIP calculator.

React Financial Data

5dashboard tabs

Architecture

The 4-Layer Memory Stack

Lumen's memory architecture is the core IP. Four layers, each solving a different retrieval problem, unified into a single context pipeline.

Layer 1 · Speed

Redis Ring Buffer

40-turn sliding window per user, cross-channel unified, compressed at 500-char boundary

Layer 2 · Relevance

Hybrid Search + RRF

BM25 + pgvector cosine → Reciprocal Rank Fusion → LLM re-ranking of top-8

Layer 3 · Relationships

Neo4j Bi-temporal Graph

Expire+create edges, full relationship history, temporal asOf queries

Layer 4 · Salience

Ebbinghaus Decay

Importance feedback loop (+0.05 per retrieval, capped 0.95). Memories that matter persist.

LLM Routing

Local-First Fallback Chain

Ollama → Anthropic → Gemini → OpenAI. Free local inference first, cost tracking per call.

Agent Framework

10 Persona Agents

Specialized agents with soul configs, composable skills, A2A protocol, multi-channel routing

Philosophy

How I Think

Architecture is the art of making the right tradeoffs visible. Good systems reveal their intent through structure — the best documentation is the code itself.

The measure of an architect isn't the complexity they can build — it's the complexity they can eliminate while preserving capability.

Design with Intent

Every layer, every interface, every naming convention should communicate why it exists. If a junior engineer can't trace a decision back to a business constraint, the architecture failed.

Govern, Then Automate

AI adoption without governance is technical debt with a marketing budget. Establish guardrails first — compliance, data lineage, access controls — then accelerate.

Simplicity is a Feature

Three well-designed microservices beat twelve tangled ones. Attention to detail in naming, spacing, error handling, and observability is what separates production systems from prototypes.

What's Next

Where I'm Going

Building at the intersection of enterprise architecture and AI product development, with a clear trajectory toward executive leadership and entrepreneurship.

Enterprise Architecture & Cloud Modernization

What I Want to Build

Large-scale cloud transformations — migrating legacy platforms to modern, governed architectures with measurable ROI
AI/ML operationalization in regulated industries — bridging the gap between proof-of-concept and production-grade deployment
Data strategy and platform architecture — building the data foundations that make AI adoption possible at enterprise scale

AI Products & Developer Tools

What I'm Exploring

Smara — building the default memory layer for AI agents — persistent context at a fraction of the cost of existing solutions
Developer tooling — MCP Doctor, Scoper, and open-source contributions to the AI developer ecosystem
Multi-agent systems — production orchestration patterns, memory architectures, and autonomous workflows

Interested in working together? I'm open to architecture & cloud leadership, functional consulting, and AI product partnerships.