anurag Contact
engineering · backend · applied AIGurugram, India

whoami

Anurag
Sharma.

Engineering Manager.
Software developer at heart.

I build backend systems, ship production AI, and lead engineering teams. 10+ years turning complex problems into reliable software, with 3+ years delivering GenAI and ML at scale.

View achievements GitHub (opens in a new tab) LinkedIn (opens in a new tab)
profile.pyPython
# A little about me
engineer = {
    "name": "Anurag Sharma",
    "role": "Engineering Manager",
    "company": "Tata 1mg",
    "experience": "10+ years",
    "production_ai": "3+ years",
    "team_size": 10,
    "promotions_enabled": 2,
    "focus": [
        "Distributed systems",
        "GenAI & ML",
        "Engineering leadership",
    ],
}
 
# Architecture → delivery → impact
Python · Backend · AI/ML
// engineering outcomesTata 1mg
Fewer escalations
50%
Customer satisfaction
+10 pp
SKUs with QR traceability
750K+
Shorter release cycles
25%
// 01 — Achievements

Selected achievements.

Selected systems I’ve built and led at Tata 1mg,
with outcomes that made a difference.

Conversational AI

15 → 7 min

Faster customer resolution

Built a GenAI-first chatbot with retrieval over order, policy, and catalogue data, plus tool calling into order and refund systems.

CSAT
+10 pp
Agent handoff
−14 pp
LLM · RAG · Tool callingUSD 0.05 / conversation
Email automation

15 → 5

Fewer agents required

Automated email replies and ticket resolution with intent classification, retrieval, drafting, and guardrails.

First-contact resolution cycle12–18 hours
Classification · Retrieval · Guardrails
Predictive ML

50%

Reduction in escalations

Trained and deployed a custom CSAT-risk model to score live conversations and identify escalation risk.

At-risk conversationsRouted to senior agents
Custom model · Live risk scoring
Retail & fulfillment

750K+

SKUs made traceable

Owned retail and fulfillment engineering, extending QR traceability across more than 750,000 SKUs.

Retail order automationTAT breaches: 26% → 9%
Order lifecycle · QR traceability
Engineering delivery

25%

Shorter release cycles

Improved CI/CD, code review, and release processes to increase engineering throughput.

Team development10 engineers · 2 promotions
CI/CD · Code review · Leadership
Security automation

Weeks→Days

Faster vulnerability fixes

Built an AI-driven suite that scans for CVE and DeepSec vulnerabilities, opens fix pull requests, and validates patches in CI/CD.

Automated workflowScan → fix PR → validation
LLM agents · CVE · CI/CD
Business impact

Better order experience. Measurable revenue.

Revamped the Order Success page as part of order lifecycle engineering.

₹1.2 lakh/dayIncremental revenue

pp = percentage points · CSAT = customer satisfaction · TAT = turnaround time

// 02 — Experience

The journey
behind the code.

From real-time platforms and distributed systems to engineering leadership and applied AI.

2016 — Present
Aug 2022 — PresentCurrent

Tata 1mg

Associate Engineering Manager

Customer Experience Platforms & GenAI Gurugram, India

Lead GenAI and ML engineering for customer-support platforms. Own architecture, roadmap, and delivery with product and operations leadership, managing a 10-member team with two promotions.

  • Built the customer-support chatbot, email automation engine, and live CSAT-risk model behind the outcomes above.
  • Built a computer-vision classifier to screen prescriptions and uploaded images for invalid, duplicate, or AI-generated content, automating review and speeding verification.
  • Created AI-driven security automation for CVE and DeepSec scanning, fix pull requests, and CI/CD validation, cutting remediation from weeks to days.
  • Established LLM evaluation, prompt versioning, quality gates, and cost/latency monitoring. Standardized a shared model gateway and retrieval layer across chatbot and email automation.
  • Improved CI/CD, code review, and release processes, reducing release cycle time by 25%.
  • Owned order lifecycle, retail, and fulfillment engineering: an Order Success revamp generated ₹1.2 lakh/day incremental revenue; retail automation lowered TAT breaches from 26% to 9%; QR traceability covered 750K+ SKUs.
LLM + RAGML systemsArchitectureTeam leadership
Apr 2021 — Jul 2022Bengaluru, India

Hubilo Softech

Lead Software Engineer

  • Designed a Kafka and Debezium change-data-capture pipeline powering a reusable event bus and search, contributing to a 10% increase in retention.
  • Built a high-volume notification engine for email, push, and WebSocket events with 10% ROI. An in-house template engine cut costs by 10%.
  • Built a rule-based CPE certification engine for virtual events, improving attendee engagement by 20–30%.
KafkaDebeziumEvent-driven systems
Apr 2019 — Mar 2021Kuala Lumpur, Malaysia

Millennium System Design

Software Engineer · SDN BHD

  • Built a real-time chat microservice from scratch with Node.js, Redis, Socket.IO, and Kafka for full-duplex communication at scale.
  • Built serverless Golang image-processing APIs with in-memory OpenGL shaders and shipped a Samsung TV streaming app with 11 camera views.
Node.jsGolangReal-time systems
Feb 2017 — Apr 2019Noida, India

TO THE NEW

Software Engineer

  • Built a unified notification service across email, push, and in-app channels, optimizing delivery latency.
  • Migrated REST services to GraphQL, cutting startup time by 30% and improving application performance.
Feb 2016 — Feb 2017Delhi, India

Stealth Startup

Software Engineer

  • Designed a data ingestion pipeline aggregating product data across e-commerce platforms and built automation and reporting utilities.
// 03 — Tech stack

Tools I build with.

Hands-on across AI, backend engineering,
and the teams bringing them together.

GenAI & ML

From retrieval and agent workflows to evaluation, deployment, and guardrails.

  • LLMs · OpenAI & Anthropic Claude APIs
  • Prompt engineering · RAG
  • Embeddings · Vector search
  • Agentic workflows · Function calling
  • LLM evaluation · Guardrails
  • NLP · Computer vision · Classification
  • scikit-learn · PyTorch
  • Model deployment · MLOps

Backend & cloud

Distributed systems, event-driven architecture, and reliable delivery.

  • Python · Java · Golang · Node.js
  • Microservices · REST APIs · GraphQL
  • Apache Kafka · RabbitMQ
  • Redis · PostgreSQL · MySQL · MongoDB
  • Distributed & event-driven systems
  • Amazon Web Services
  • Docker · CI/CD
  • Vulnerability management · CVE

Engineering leadership

Technical direction, team growth, and execution with business partners.

  • Engineering management
  • Team leadership · 10 engineers
  • Career development · 2 promotions
  • System architecture
  • Roadmap ownership
  • Stakeholder management
  • Agile delivery
// 04 — Education

The foundations.

Aug 2014 — Jul 2017

Master of Computer Applications

Guru Gobind Singh Indraprastha University

Exemplary Student Award
Jul 2010 — Jul 2013

B.Sc. (Honours) in Physics

University of Delhi