Anurag Kumar Naik — Director of Engineering

𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 𝐬𝐜𝐚𝐥𝐚𝐛𝐥𝐞 𝐀𝐈 𝐬𝐲𝐬𝐭𝐞𝐦𝐬 | 𝐆𝐞𝐧𝐀𝐈, 𝐀𝐠𝐞𝐧𝐭𝐬, 𝐌𝐋, 𝐃𝐚𝐭𝐚 𝐏𝐥𝐚𝐭𝐟𝐨𝐫𝐦𝐬, 𝐁𝐚𝐜𝐤𝐞𝐧𝐝 𝐚𝐧𝐝 𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐞𝐝 𝐒𝐲𝐬𝐭𝐞𝐦𝐬 I'm a problem-solver at heart — passionate about building intelligent systems that scale. With over 8 years of hands-on experience across AI engineering, backend systems, data platforms, and distributed computing, I’ve had the opportunity to build and lead impactful products that blend innovation with real-world utility. 𝐀𝐈 & 𝐆𝐞𝐧𝐀𝐈 𝐚𝐭 𝐒𝐜𝐚𝐥𝐞 At INDmoney, over the past 3+ years, I've built 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧-𝐠𝐫𝐚𝐝𝐞 𝐀𝐈 𝐬𝐲𝐬𝐭𝐞𝐦𝐬, from GenAI-powered content generation (INDnews) to a 𝐦𝐮𝐥𝐭𝐢-𝐚𝐠𝐞𝐧𝐭 𝐀𝐈 𝐭𝐫𝐚𝐝𝐢𝐧𝐠 / 𝐢𝐧𝐯𝐞𝐬𝐭𝐢𝐧𝐠 𝐚𝐬𝐬𝐢𝐬𝐭𝐚𝐧𝐭 (mIND) using LangGraph — enabling users to screen stocks, indices, and F&O with natural language across technical and fundamental signals. 𝐌𝐚𝐜𝐡𝐢𝐧𝐞 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 & 𝐈𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐨𝐧 With 5+ years in ML, I’ve developed systems like a high-reliability PDF Parsing Engine powered by a custom DSL that enables creating document parsers in 15–20 minutes — scalable and adaptable across formats. This is heart of asset tracking at INDmoney. 𝐀𝐩𝐩𝐥𝐢𝐞𝐝 𝐀𝐈/𝐌𝐋 𝐟𝐨𝐫 𝐁𝐮𝐬𝐢𝐧𝐞𝐬𝐬 𝐈𝐦𝐩𝐚𝐜𝐭 At Goibibo, I architected a dynamic pricing engine that adjusts hotel prices hourly based on margin targets — delivering 30% more bookings at reduced discounts. This is core of Hotel Price at Goibibo. 𝐑𝐨𝐛𝐮𝐬𝐭 𝐁𝐚𝐜𝐤𝐞𝐧𝐝 𝐒𝐲𝐬𝐭𝐞𝐦𝐬 I’ve built modular, scalable 𝐛𝐚𝐜𝐤𝐞𝐧𝐝 platforms, including a lending system integrating with 5+ partners for both secured (InstaPlus - Loan Against Mutual Funds) and unsecured credit (InstaCash - Credit Line) — optimized for reliability and extensibility. These are revenue powerhouses. 𝐃𝐚𝐭𝐚 𝐏𝐥𝐚𝐭𝐟𝐨𝐫𝐦𝐬 & 𝐈𝐧𝐟𝐫𝐚 From building GoMemory at 𝐆𝐨𝐢𝐛𝐢𝐛𝐨, streaming engine that stores user's activity footprints to personalize experience, portfolio tracking systems at 𝐈𝐍𝐃𝐦𝐨𝐧𝐞𝐲 (credit cards, stocks, etc.), to contributing to the Data Platform for 𝐀𝐚𝐫𝐨𝐠𝐲𝐚 𝐒𝐞𝐭𝐮 (𝐆𝐨𝐯𝐭. 𝐨𝐟 𝐈𝐧𝐝𝐢𝐚) — I’ve worked on systems that handle scale, compliance, and real-time performance

Stackforce AI infers this person is a Fintech and Data Engineering expert with strong AI capabilities.

Location: Bengaluru, Karnataka, India

Experience: 8 yrs 11 mos

Skills

Ai Engineering
Multi-agent Systems
Software Engineering
Machine Learning
System Architecture
Content Generation
Product Management
Data Engineering
System Design
Data Analysis

Career Highlights

Over 8 years of experience in AI engineering.
Developed production-grade AI systems for fintech.
Led impactful projects with significant user engagement.

Work Experience

INDmoney

Director of Engineering (2 yrs)

Senior Software Engineering Manager (1 yr 2 mos)

Engineering Manager (1 yr 11 mos)

Principal Software Engineer (9 mos)

Aarogya Setu

Senior Data Engineer (Volunteer) (3 mos)

Goibibo

Senior Software Engineer I (1 yr 6 mos)

Software Engineer (1 yr 9 mos)

I4U Labs, Bangalore.

Android Developer (2 mos)

Education

Bachelor’s Degree at National Institute of Technology Calicut

12th at Govt. Boys Higher Secondary School Baloda

Anurag Kumar Naik

Director of Engineering

Bengaluru, Karnataka, India8 yrs 11 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Over 8 years of experience in AI engineering.
Developed production-grade AI systems for fintech.
Led impactful projects with significant user engagement.

Stackforce AI infers this person is a Fintech and Data Engineering expert with strong AI capabilities.

Contact

Skills

Core Skills

Ai EngineeringMulti-agent SystemsSoftware EngineeringMachine LearningSystem ArchitectureContent GenerationProduct ManagementData EngineeringSystem DesignData Analysis

Other Skills

AI AgentsAPI GatewayAPI GatewaysAWS LambdaAirflowAmazon RedshiftAmazon SQSAmazon Web Services (AWS)Analytical SkillsAndroid DevelopmentApache AirflowChatGPTDebeziumDefining RequirementsDelta.io

About

Experience

8 yrs 11 mos

Total Experience

4 yrs 5 mos

Average Tenure

5 yrs 8 mos

Current Experience

Indmoney

4 roles

Director of Engineering

Promoted

Jun 2024 – Present · 2 yrs

Trading AI Assistant (Mind):
Architect-ed and coded hands on "a powerful Multi AI Agent System" that helps user to do Technical Analysis (using realtime indicators and OHLCV data across all Indian Stocks, screening based on any criteria).
Option Chain Analysis to help understand where the opportunity lies, support resistance, strike price analysis, greeks analysis all explained in easy to understand language.
Fundamental Screening, analysis or screen stocks using company financials data (Quarterly results, , Annual Finances, Shareholding Patterns etc) all in just seconds!
Built using Multi-Agent AI framework LangGraph, Langchain, and RAG techniques.
Insta Plus: Loan Against Mutual Funds
Built LaMF from scratch contributing to big revenue chunk.
Chart Pattern Detection: INDstocks chart pattern detection realtime and sending nudge to traders to help make decision and save time in searching for patterns on each stocks! Patterns like Channel, Wedge, Symmetric Triangle, Flag, Pennant, Cup & Handle, Head & Shoulder etc are implemented.

AI AgentsMulti-agent SystemsGenerative AIRetrieval-Augmented Generation (RAG)AI Engineering

Senior Software Engineering Manager

Apr 2023 – Jun 2024 · 1 yr 2 mos

INstaCash:
☑ Created Lending Stack from scratch and done disbursement of ~1200CR+
☑ Led the team and architected the entire system which enabled integration with 4 different lending partners.
☑ Contributed with UX Design and constant Product inputs.
☑ The tech enables sending contextual communications to the user base, interactions with partners, and CX admin panel for faster dispute resolutions.
Tech Stack: GoLang, API Gateway, Lambda, SQS, Postgresql, Newrelic, Loki + Grafana.
INDstories: News Content Generation & Nudges (Generative AI Technology):
☑ Created a News Summarisation Platform and Product from Scratch, with Generative-AI using GPT.
☑ The system does news aggregation and summarises it in a very crisp language, creates timeline of the events and summary for the day.
☑ Created a trending stocks algorithm based on price, volume and news inflow.
☑ Product explains, "Why stock price has moved up / down using the news out there for the day".
☑ The content is generated on a semi-realtime basis.
☑ System creates Nudges, a very short one liner on some significant news for the stock.
☑ Project has delivered amazing user engagement and retention.
Tech Stack: Python, chat-gpt-3.5-turbo, Lambda, API Gateway, SQS, Postgresql, Redshift, Redash.

High-Level DesignEngineering ProcessAmazon Web Services (AWS)PythonPostgreSQLAmazon SQS+16

Engineering Manager

Promoted

Jul 2021 – Jun 2023 · 1 yr 11 mos

𝐈𝐏𝐎 𝐂𝐞𝐧𝐭𝐞𝐫:
☑ This project was built in record 3 days time from ideation to the execution end to end, as ZOMATO's IPO was just 4 days ahead.
☑ It is the fastest executed project at INDMoney!
☑ Led & created an easy place to apply for an IPO in just 3 simple clicks!
☑ Challenge: In IPO application, one needs to fill the broker information which has a unique set of identifiers for every user, DPID and ClientId (sometimes referred as BOID, both combined). Finding these ids from your broker websites may vary and typically it is extremely hard to find these information given there are numerous stock brokers in India.
☑ Since we track all of user's investments across broker, IPO center pre-fills these Ids for you!
☑ User can see upcoming IPOs, Closed IPOs and their application status at one place.
☑ Gets real-time update on the application as soon as it is changed at exchange and communicates the user on application progress via emails/notifications.
𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝: Scala, Play Framework, Postgres, Amazon SQS, Python, Airflow.
𝐂𝐫𝐞𝐝𝐢𝐭 𝐂𝐚𝐫𝐝 𝐓𝐫𝐚𝐜𝐤𝐢𝐧𝐠 𝐒𝐲𝐬𝐭𝐞𝐦:
☑ Led, created architecture and built the Credit Card Tracking System from scratch.
☑ Created CC "mini-app" as a separate vertical on INDMoney App.
☑ Created Data Pipelines to read documents from GMAIL, parse it using INDvision (the PDF parsing engine I built earlier), we have more than 400 Credit Card's Variant Parser.
☑ System rewards user with US Stocks for spending on selected US brands.
☑ The mini-app serves the analytics on the due information, spend on merchants, and historical statements details etc.
☑ Sends timely communications on new statement's arrival and reminder to pay bills.
𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝: Scala, Play Framework, Postgres, Amazon SQS, Python, SQL, Airflow.

High-Level DesignLow-Level DesignProblem SolvingDefining RequirementsNew Product RolloutProduct Management+1

Principal Software Engineer

Oct 2020 – Jul 2021 · 9 mos

𝐈𝐍𝐃𝐯𝐢𝐬𝐢𝐨𝐧, 𝐚 𝐆𝐞𝐧𝐞𝐫𝐢𝐜 𝐏𝐃𝐅 𝐏𝐚𝐫𝐬𝐢𝐧𝐠 𝐄𝐧𝐠𝐢𝐧𝐞:
☑ Designed and implemented a generic PDF parsing platform from the scratch. Platform is responsible to extract "any" information out of PDF document
☑ Created own DSL, which any non-technical person can use and extract the information
☑ System can extract Text, Tables (Detection and extraction for any type of table) and Components (combination of the two with information tied together)
☑ System is built to solve for 100% reliable extraction and go-to-market time for any new document within less than an hour
☑ Uses spatial information to extract the data (that's why the name "vision") rather than relying on REGEX, this provides edge in accuracy
☑ Currently getting used to parse all the PDF documents like Mutual Fund AMC and CAS statements, EPF, NSDL, CDSL holding, NPS document etc
☑ Utilized to validate KYC documents by parsing ITR, Indian Bank Statements etc for remittence
☑ Credit Card Statement parser for more than 600 CC Statement Variants
𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝: Python, Flask, PDF libraries.
𝐈𝐧𝐝𝐢𝐚𝐧 𝐒𝐭𝐨𝐜𝐤𝐬 𝐓𝐫𝐚𝐜𝐤𝐢𝐧𝐠 𝐒𝐲𝐬𝐭𝐞𝐦:
☑ One of the complex and challenging problem that I have led and solved so far, Indian Stocks Tracking
☑ Created a new parallel system to track Indian stocks, to replace older, migrated all the users to new system reliably
☑ Challenge: In India, there is no single source available to provide your stocks information across all the brokers.
☑ System can track stock holdings irrespective of the number of brokers you are invested in.
☑ Built a system which reads document from user's emails and gets Depository and Exchange statements and parses using INDvision to stores in data store. It intelligently does the de-duplication and merges the holdings + trades data construct final holdings of user.
☑ Created reliable and faster APIs to serve portfolio information.
𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝: Scala Play, Postgres, SQS, Python, Airflow.

High-Level DesignLow-Level DesignProblem SolvingDefining RequirementsNew Product RolloutSoftware Engineering+1

Aarogya setu

Senior Data Engineer (Volunteer)

Apr 2020 – Jul 2020 · 3 mos · Bengaluru, Karnataka, India

Built the privacy-first data-platform and insight platform for India's contact tracing app.
☑ Built the first-version of Data Platform for AarogyaSetu, along with a small team in a couple of days.
☑ Wrote MySQL and DynamoDB to Redshift pipelines.
☑ Setup Debezium and Airflow for real-time ingestion and pipelines.
☑ Built and maintained ICMR Data to Data Warehouse pipeline. This data is very crucial for AarogyaSetu as this is the source for COVID-positive users, recovered users, and users who tested negative.
☑ Built a pipeline to forecast positive cases across states using fb-prophet.
☑ Modified Open Source Spark-DynamoDB connector to improve write speed by 10x (https://github.com/AnuragNaik/spark-dynamodb).
𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝: Spark, Scala, Kafka, Debezium, Delta.io, Airflow, ElasticSearch, and Redash.

Analytical SkillsData Engineering

Goibibo

2 roles

Senior Software Engineer I

Promoted

Apr 2019 – Oct 2020 · 1 yr 6 mos

𝐃𝐲𝐧𝐚𝐦𝐢𝐜 𝐃𝐢𝐬𝐜𝐨𝐮𝐧𝐭𝐢𝐧𝐠 (June 2018-Oct 2020):
☑ Designed and programmed a near-real-time, reinforcement learning-based, scalable, and "Fully Automated Dynamic Discounting" for Hotels at Goibibo from scratch.
☑ Impact: It gives >20%-30% more business while giving 10%-40% lesser discounts.
☑ The only input it takes is profit percentage and internally maximizes transactions by setting discounts for each hotel@checkin combinations.
☑ By doing continuous AB, it learns price elasticity of demand at runtime.
☑ Under the hood, it combines entities with similar price elasticity, does transaction forecasting and meets the business criteria set by the revenue team.
☑ Built a custom algorithm for clustering hotel@checkin, inspired by k-d tree and theory of information gain.
☑ Built hierarchical optimizers that take business constraints for individual leaf nodes and the top level. It satisfies all the constraints and maximizes transactions.
☑ Implemented the distributed computing component in a high-performance Scala-Spark framework.
☑ Implemented Machine Learning components of the algorithm in the Python-Pandas framework.
☑ Connected Scala-Spark with Python through gRPC by using the Sidecar Pattern.
☑ Added multivariate A-B testing support. This allowed us to modify/introduce a new module in the Engine and choose the best (Control vs Test) performing variant by looking at the result.
☑ Built an Expectation Checker framework for Spark from scratch to validate the discounts before it gets applied.
☑ Used property-based testing to validate the correctness of the core components of the project.
☑ Created Dashboards on Redash for Operational Analytics and Alerting.
☑ Worked as a tech lead, architect, and a developer. Did quarterly planning and sprint execution. Managed product and business stakeholders.
𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝: Spark, Scala, Delta.io, Python, Pandas, gRPC, SQL, Redshift, Redash, Airflow.

Analytical SkillsProblem SolvingSoftware EngineeringData Engineering

Software Engineer

Jun 2017 – Mar 2019 · 1 yr 9 mos

𝐆𝐨𝐌𝐞𝐦𝐨𝐫𝐲:
☑ GoMemory, as the name suggests, it is Goibibo's Memory of a user. It is about having all user-data at one single place.
☑ It acts as an event-store for backends and model serving applications. Provides the requested events for a user within < 10ms latency.
☑ A context initializer for GIA (Goibibo Support ChatBot). Helps in showing relevant options proactively even before the user starts the conversation.
☑ It ingests data from 20+ different event streams eg. transaction, web-footprints, wallet, ratings & reviews, etc with bounded SLA (< 50ms).
☑ Built Code Generator, an SBT-Plugin to convert events in protobuf format to GoMemory format. This zero downs go-to-production time.
𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝: Spark, Scala, Kafka(Kafka Streams + Kafka Connect), DynamoDB, Redis, Python, Play Framework's Twirl-Template-Engine.

I4u labs, bangalore.

Android Developer

May 2016 – Jul 2016 · 2 mos · Bengaluru Area, India

☑ A GPS based navigation application (Android) using OsmAnd Framework, coded in Java.
☑ It was an early-stage startup, my role was feature enhancement and fixing issues.
☑ App didn't had download recovery for Map files from the last checkpoint. In case of download failure, the app used to download the entire file of > 500MB from the beginning. Added resume capability for the client in case of incomplete download and many behavioural fixes.