ANKUR

15+ YEARS IN PRODUCTION

Senior Engineering Leader — AI Systems, Platform Architecture & Fintech Engineering

Chennai, India | ankur.cse.bvcoe@gmail.com | +91 9971306758 | linkedin.com/in/codingmentor

1,000+ ENGINEERS ENABLED 54% AI COST REDUCTION $4.28M DELIVERED 25+ ENGINEERS LED

01Professional Summary

Senior Engineering Leader with 15+ years designing and scaling distributed, high-availability fintech platforms, now specializing in AI-native engineering leadership — architecting and operating production-grade, self-hosted LLM systems in regulated financial environments. Currently own engineering for IDFC FIRST Bank's Wealth Management platform (Stocks, Mutual Funds, IPOs, Sovereign Gold Bonds, Government Securities, Insurance) and Financial Management, spanning 80+ microservices. Designed and led idfc-coder, a self-hosted AI code-generation platform (AWS SageMaker, DeepSeek Coder, Mistral) serving 1,000+ engineers, replacing third-party Copilot/Claude API spend and cutting AI tooling cost by 54%. Prior distributed-systems experience across PayPal's global settlement infrastructure, TomTom's real-time geospatial data pipeline, and BlackRock's credit-risk platform. Track record building and mentoring engineering teams of 25+; former startup founder (EdTech). Targeting an AI Leadership role combining hands-on platform architecture with organizational AI strategy.

02Technical Skills

AI / ML Systems: Agentic AI architecture, self-hosted LLM deployment (DeepSeek Coder, Mistral, WizardCoder, Llama), model evaluation & benchmarking, RAG (Retrieval-Augmented Generation), vector databases, prompt engineering, GenAI/LLM API integration (Claude, GPT), AI safety & governance for regulated systems, AI-assisted SDLC, canary model deployment, inference optimization (quantization, batching, latency tuning).

Languages & Frameworks: Java, Spring Boot, REST APIs, SQL, Python (scripting/tooling).

Cloud & Infrastructure: AWS (SageMaker, ECS Fargate, Lambda, ALB, CloudWatch, X-Ray), GCP, serverless architecture, Kubernetes-adjacent container orchestration, Infrastructure-as-Code, CI/CD pipelines, DevSecOps.

Distributed Systems: Microservices architecture, event-driven systems (Apache Kafka), pub/sub messaging, SQS-based queuing/batching, circuit breaker & resilience patterns, canary deployments with automated rollback, high-throughput/low-latency system design.

Data & Caching: SQL, NoSQL (MongoDB), Redis / ElastiCache, GemFire, DynamoDB, batch and streaming data pipelines, cache synchronization & serialization.

Fintech Domains: Core banking, payments & settlement, wealth management (equities, mutual funds, IPO, sovereign gold bonds, government securities, insurance), personal finance management (PFM), Account Aggregator framework, regulatory compliance, credit risk, fraud detection.

Engineering Leadership: Team scaling and mentorship (25+ engineers), technical roadmap ownership, OKR-driven delivery, cross-functional stakeholder management (Product, Risk, Compliance, Security), architecture review, hiring.

03Flagship AI Initiative: idfc-coder — Self-Hosted AI Engineering Platform

IDFC FIRST Bank | Conceived, architected, and led end-to-end delivery

Designed and shipped a self-hosted, VPC-only AI code-generation platform to replace third-party Copilot/Claude API spend across the engineering organization, built for data sovereignty, cost control, and production reliability under banking-grade compliance requirements.

Model selection & evaluation: Benchmarked 8 candidate open-weight LLMs (DeepSeek Coder 33B, Mistral 7B, WizardCoder 34B, CodeLlama 34B, Llama 2 70B, StarCoder2, GPT-NeoX, Phi-2) against 500 real internal pull requests; selected a multi-model serving strategy — DeepSeek Coder 33B as primary, Mistral 7B as a low-latency fast path.
Infrastructure architecture: Deployed models on AWS SageMaker (ml.p4d.24xlarge, A100 GPUs) behind an ECS Fargate wrapper service and ALB, with Redis/ElastiCache L1 caching (40% hit rate), DynamoDB L2 caching, and SQS-based request batching; configured auto-scaling across 4–16 instances.
Reliability engineering: Implemented circuit breaker patterns (closed/open/half-open), canary deployment pipelines (10% → 25% → 50% → 100% rollout with automated rollback triggers), and full observability via CloudWatch and X-Ray.
AI governance & safety: Designed a code-criticality tiering model (liability tiers) restricting AI-generated code in payment- and compliance-adjacent systems, with mandatory human-verification gates and audit-trail logging for regulatory review.
Workflow automation: Built a Jira-integrated agent that polls in-flight tickets, generates and validates candidate code, and opens pull requests automatically — reducing manual hand-off time across squads.
Performance & adoption results: Reduced inference p95 latency from 3.5s to 620ms (via request batching, streaming responses, and INT8 quantization); delivered approximately $4.28M in savings within the first 6 months (54% cost reduction, $7.9M → $3.6M/yr run rate); achieved 68% suggestion-acceptance rate, +42 developer NPS, and a 40% increase in sprint velocity across a phased rollout from 100 to 1,000 engineers.

Technical stack: AWS SageMaker, ECS Fargate, Application Load Balancer, Kafka, SQS, Redis/ElastiCache, DynamoDB, CloudWatch, X-Ray, DeepSeek Coder, Mistral, WizardCoder, Claude.ai, Jira REST API, CI/CD with canary rollout and automated rollback.

04Professional Experience

Senior Engineering Manager — IDFC FIRST Bank

Chennai, India | May 2025 – Present

Own engineering for the Wealth Management platform (Stocks, Mutual Funds, IPOs, Sovereign Gold Bonds, Government Securities, Insurance) and Financial Management, spanning 80+ microservices.

Lead a team of 25+ engineers and tech leads across Financial Management and Wealth Management, owning architecture direction and AI-adoption strategy.
Conceived, architected, and delivered idfc-coder, IDFC's self-hosted AI coding platform — full ownership of model selection, infrastructure, safety governance, and production rollout (see Flagship AI Initiative above).
Led cloud-native modernization of the platform — microservices decomposition, event-driven architecture on Kafka, and DevSecOps pipeline hardening — while embedding AI-assisted workflows into the SDLC to improve scalability and developer velocity.
Partner directly with Product, Business, Risk, Compliance, and Security to ship regulatory-compliant, AI-assisted features on a predictable delivery cadence.

Engineering Manager — IDFC FIRST Bank

Chennai, India | May 2024 – April 2025

Managed a 14-person PFM squad (8 developers, 2 QA, 2 UAT, 1 BA, 1 Tech Lead) delivering core features for India's #1 Forrester-rated mobile banking app.
Shipped Account Aggregator–based Consolidated Bank Statements, enabling secure, consent-driven, multi-institution financial data aggregation at scale.
Architected and launched the Net Worth platform — a unified view of customer assets (EPF, stocks, mutual funds, fixed deposits) and liabilities (loans, credit cards, mortgages) — driving measurable increases in engagement and retention.
Standardized engineering quality gates (CI/CD, automated testing, SLA dashboards), improving release velocity by 30% and reducing production incidents.

Member of Technical Staff — PayPal

Bangalore, India | November 2021 – April 2024

PayPal's global settlement and payments infrastructure, processing billions of dollars in daily transaction volume.

Designed and built Java-based microservices and pub/sub consumers within PayPal's core settlement system, ensuring accurate, real-time financial reconciliation at global scale with strict consistency guarantees.
Co-architected a generic case-management and reporting platform adopted across multiple compliance and operational workflows, significantly reducing manual investigation effort.
Contributed to the fraud-detection framework — designing data models and rule-based detection logic for identifying and reporting suspicious transactions across global payment flows.

Technical Lead — TomTom

Bangalore, India | March 2019 – October 2021

Location technology and map data — Points of Interest (POI) data pipeline serving millions of connected devices.

Architected a serverless cloud-export pipeline (AWS Lambda-based) that cut export processing time exponentially, unlocking near real-time data delivery for downstream consumers.
Introduced fully automated one-click build and release pipelines, eliminating manual deployment steps and cutting release cycle time by approximately 70%.
Owned end-to-end architecture and delivery for three core microservices in the POI ingestion system; led design reviews and mentored a team of engineers.

Product & Engineering Lead / Founder — Coding Ninjas & CodingS9

Delhi, India | May 2016 – March 2019

Joined as the 2nd engineer at Coding Ninjas (now a multi-million-dollar EdTech company), contributing to platform architecture during its formative growth phase.
Founded CodingS9, an EdTech startup — led product design, engineering, business development, and hiring from zero to launch.
Built the CodeZen online coding platform end to end, owning the full product lifecycle from wireframe to production.

Assistant Manager – Engineering — Citicorp & BlackRock

India | May 2014 – May 2016

Citicorp: global banking and financial services (core banking, payments, institutional finance). BlackRock: credit-risk platform processing millions of daily transactions across global portfolios.

Resolved a critical GemFire cache-synchronization bug using custom serializers, exponentially reducing sync latency and improving reliability for credit-risk calculations.
Optimized high-frequency batch processes using JProfiler-based performance analysis, achieving a 60% improvement in throughput and directly reducing end-of-day risk-computation time.
Built and maintained banking technology platforms supporting high-volume transaction processing across retail and institutional business lines, ensuring regulatory compliance and high availability.

Senior Software Engineer — Samsung Research India

Bangalore, India | September 2010 – April 2014

Implemented SEAndroid security policies for commercial flagship devices (Galaxy Note 3, Galaxy S3, Galaxy S4 Mini), contributing to enterprise-grade device hardening.
Engineered new ActiveSync email capabilities for commercial smartphone lines, shipped across millions of devices globally.
Developed the Mobile Device Management (MDM) module enabling over-the-air enterprise management of Android devices.
Selected as team representative for on-site project commercialization in South Korea, recognized for technical ownership and cross-functional collaboration.

05Key Achievements

AI platform leadership: Conceived and led idfc-coder, IDFC's self-hosted AI coding platform for 1,000+ engineers — evaluating, deploying, and governing self-hosted LLMs in a regulated banking environment; established as the bank's internal benchmark for AI-augmented engineering delivery.
Product leadership: Contributed to the mobile banking app rated #1 in India by Forrester for three consecutive years (Financial Management & Wealth Management).
Startup to scale: Early-stage engineer at Coding Ninjas (2nd hire) — a startup that scaled to a multi-million-dollar EdTech business.
Global engineering exposure: Shipped code on flagship Samsung devices distributed to millions of users globally; represented the team on-site in South Korea for commercialization.

06Education

B.Tech / B.E., Computer Science & Engineering

Bharatiya Vidyapeeth College of Engineering, IP University, India | 2006 – 2010