Apr 2024 – Sep 2025 · 1 yr 5 mos · Bengaluru, Karnataka, India

• Personalisation Platforms, MarTech and Agentic AI Architecture

May 2022 – Apr 2024 · 1 yr 11 mos · Bengaluru, Karnataka, India

• Optimising customer experience by building a brand that learns and grows with you.

Apr 2020 – Jun 2020 · 2 mos

Apr 2018 – Apr 2020 · 2 yrs

• Trip ideas - The travel inspiration platform of MakeMyTrip. Worked on content ingestion, discovery, ranking, recommendation of feed and destinations. Feedback is implemented in serverless architecture on AWS. Technologies used : Vert.x, Redis, Kafka, MySQL, Kinesis Stream, AWS lambda, DynamoDB
• Knock-Knock - India's first ever live travel game show. This was live show with live streaming where users could bid on packages, play quizzes etc live on the app. Worked on the backend side for this. Technologies used : Vert.x, SSE(Server sent events), Redis
• Spider - The user engagement program of MakeMyTrip. This is entirely event driven service which listens to various user events like app install, drop-off, booking etc and create relevant engagement loops for the users to increase user interaction with our app. Technologies used : Vert.x, Kafka, RabbitMQ, Redis, MySQL.

Sep 2016 – Mar 2018 · 1 yr 6 mos

• Peitho - The persuasion engine of MakeMyTrip.
• This can target the user at any/all funnel steps. It's a compute and data heavy service as the insight data is huge and hundreds of persuasions needs to be computed for hundreds of nodes(flights, hotels etc) in each request. To solve this, the service is designed as a distributed system which deterministically distributes insight data and compute tasks.
• Tech: Vert.x, Couchbase, Caffeine, Stickler.
• Stickler - It’s a jar to evaluate rules where a rule is defined as and/or of expression.
• R = ((e1&(e2|e3))&((e4&(e6|(7&e8)))|(e9|(e10&e11))))
• Expression = lhs operator rhs
• Lhs and rhs can again be complex expressions.
• Each rule is considered as graph and a forest of all graphs are created.
• It has its own scheduler which decides to run a node in queue (same strand) or a new strand.
• UserService - This keeps entire user data for MakeMyTrip.
• Replacing legacy system and migrating all users to new system. Supporting multiple user profiles(Business/Personal).
• Enabling mobile based login.
• It is READ optimized secured architecture supported with multi level insync-caching.
• The new user service is supported by ASK-WHAT-YOU-WANT inhouse-built query framework, which helped reducing
• >100 APIs(existing) to just 4 APIS(crud).
• We created a framework which supports
• 1. JSON based query language
• 2. CRUD operations supporting partial UPDATES and FETCHES.
• 3. same query structure / syntax for both fetch and update.
• Technologies Used : Dropwizard, Couchbase, MySQL, Kafka, Guava Cache, Kafka mirror
• Common discounting framework - This serves all lobs for defining, recommending, validating coupons and vouchers. It
• supports lob agnostic, bulk, generic and instant coupons with multiple validation criteria, usage restrictions etc. All
• coupons in MakeMyTrip are defined and maintained here. It is being used to recommend coupons for all lob listing pages.
• Technologies used : Dropwizard, Couchbase, MySQL, SQLite

Jun 2020 – Aug 2022 · 2 yrs 2 mos · Bengaluru, Karnataka, India

• Co-Founder and CTO at Mitron TV Private Limited. Focused on creating an Indian short video platform.

Jul 2011 – Sep 2016 · 5 yrs 2 mos · Bengaluru Area, India

• ODI

May 2010 – Jul 2010 · 2 mos

• Inspection Module for Core Banking Solution(CBS)