Portfolio
Selected Work
Case studies from enterprise data, GenAI, FinTech, and high-scale commerce platforms.
Case Study 01 — Deloitte · 2022–2026
Deloitte AI Assist — GenAI Copilot for Enterprise FinTech
Deployed across 35 tier-1 clients · 200% growth in platform engagement · 10–15x faster retrieval velocity
The Problem: Enterprise FinTech clients were operating on fragmented legacy systems — 9 disconnected applications, no unified data layer, manual compliance reviews taking 2–3 days per workflow.
What I Built: A GenAI copilot with hybrid RAG pipelines, semantic retrieval, and multi-agent orchestration. Embedded KYC/AML compliance filters directly into the retrieval layer with human-in-the-loop escalation.
The Hard Part: The technical build took 3 months. Getting compliance and legal teams to trust the system took 6. In regulated AI, trust is not a launch checklist — it is part of the product.
Clients
35+
Engagement Growth
200%
Retrieval Velocity
10–15x
Service Efficiency
20% → 70%
Case Study 02 — Deloitte · 2022–2026
Enterprise Data Unification — 9 Silos to One AI-Ready Architecture
38,000+ lines modernized · 90 initiatives · Zero big-bang cutover
The Problem: A major enterprise FinTech environment was running 9 disconnected legacy applications — each with its own schema, access controls, and data definitions.
What I Built: End-to-end data platform modernization spanning 38,000+ lines of code across 9 applications. Semantic bridges built domain by domain — metadata standards, hybrid retrieval architecture, compliance-aware API contracts.
The Hard Part: Leadership wanted a full Data Lake migration. I pushed back — a full migration would have taken 3 years and delivered a Data Swamp. The incremental approach delivered results within the first quarter.
Systems Unified
9 → 1
Lines Modernized
38k+
Delivery Velocity
10–15x
Business Disruption
Zero
Case Study 03 — Shiprocket · 2021–2022
Seller Intelligence Platform — Scaling Logistics for 100,000+ SME Merchants
100k+ active sellers · Multi-carrier orchestration · Real-time rate & routing engine
The Problem: Shiprocket’s SME sellers were making shipping decisions blind — no visibility into carrier performance by pin code, no predictive delay signals across 17+ courier partners.
What I Built: A seller intelligence layer surfacing real-time rate comparisons, pin-code-level delivery performance scores, and predictive delay flags. Established a shared data contract across all courier integrations.
The Hard Part: Carrier data quality was inconsistent. Solved through tiered SLA agreements and automated quality scoring surfaced back to carriers in their own dashboards.
Active Sellers
100k+
Carrier Partners
17+
Overhead Saved
~18%
Recognition
Top Impactor Award
Deloitte
Recognized for platform impact across 35+ enterprise FinTech clients
Spot Award
Deloitte
For delivering zero-disruption data migration ahead of schedule
Google Summer of Code
UNESCO · Ushahidi · 2018 & 2019
Selected twice for open-source contributions in civic tech
Ruchi Parijat — Senior Product Manager | AI, Data & FinTech Platforms 