Skip to content
Building the future of knowledge networks

Amey Patil

Co-Founder & Tech Lead @ WizzMe · AI & LLM Expert · Published Researcher

Scroll
Currently Building

WizzMe

AI-Native Knowledge Network

Transforming how people make decisions by connecting them with lived experiences from their network.

Knowledge isn't information. Knowledge is lived experience.

Consume

Get Contextual Answers

Describe your problem → Chat Agent asks clarifying questions → Searches relevant stories from people who've lived through it → Synthesizes contextual answers with sources.

Share

Document Experiences

Podcast Agent makes sharing frictionless — respond naturally via voice or text → Agent synthesizes into stories → See immediate impact as your story helps others.

Connect

Mediated Introductions

Demand capture reveals what people need → Pattern clustering motivates mentors → Contextual DMs with full context replace cold outreach.

Grow

Build Expertise

Stories accumulate showing expertise depth → Gamification with impact metrics → Trust and authority compound over time as you help more people.

Co-Founder & Tech Lead

Technical architect and execution engine — designing scalable social graph architecture, leading AI/ML systems, and driving product from zero to one.

Chat Agent — Conversational discovery + synthesis enginePodcast Agent — Voice-first frictionless experience captureThought Partner — AI companion with transparent memory retrievalSocial Graph Architecture — Scalable knowledge network backbone
// About

AtechnologistobsessedwithsolvinghardAIproblems

Over 5+ years at Flipkart, I built AI systems that serve hundreds of millions of users — from a machine translation platform spanning 11 Indian languages to FK-GPT, where the 8B model outperformed GPT-4o-mini and the 70B achieved parity with GPT-4o on e-commerce tasks at 3x lower cost. Now, as Co-Founder & Tech Lead of WizzMe, I'm applying first-principles thinking to reinvent how people access trusted knowledge from their networks.

My research has been published at ACL, EMNLP, and NAACL — the top venues in natural language processing. Before all this, I earned JEE Advanced AIR 65 and studied Computer Science at IIT Bombay.

0+

Years in AI/ML

0

JEE Adv. AIR

0

Top-Tier Publications

0x

LLM Cost Savings

// Experience

CareerJourney

From IIT Bombay to building AI systems at scale — each role building on the last.

WizzMeDec 2025 — Present
Co-Founder & Tech LeadDec 2025 — Present · Bangalore
Core Platform Architecture
Thought Partner AI
Social GraphAI SynthesisVoice-FirstPersonalization
FlipkartAug 2020 — Jan 2026 · 5 yrs 6 mos
Senior Data ScientistFeb 2025 — Jan 2026
Conversational Search Agent Search-2.0
Search Retrieval Evaluation
Agentic SearchGRPOSFTNDCG
Data Scientist — IIIOct 2023 — Feb 2025 · 1 yr 5 mos
FK-GPT Foundational LLM

8B outperformed GPT-4o-mini; 70B achieved GPT-4o parity on e-commerce. 3x cost savings vs commercial LLMs.

GenAI Chatbot Flippi
FK-GPTRLHFRanked DPOFlippi
Data Scientist — IIMar 2022 — Oct 2023 · 1 yr 8 mos
Domain-Specific BERT Models
NLP Platform
BERTMultilingual XLMNERVoice
Data Scientist — IAug 2020 — Mar 2022 · 1 yr 8 mos
Machine Translation Platform
TranslationActive LearningSynthetic DataEMNLP
Samsung R&DMay — Jul 2019
Research InternMay — Jul 2019 · Bangalore

Neural Speech-to-Speech Translation. Implemented LSTM encoder + attention-based decoder combining speech recognition, translation, and synthesis.

SpeechNMTLSTM
LTTS — IIT BombayJul 2019 — Jul 2020
Research CollaboratorJul 2019 — Jul 2020 · Mumbai

Chest X-Ray diagnosis using deep learning. Hierarchical multi-label classification with CNNs. Top 30 placement in Stanford's CheXpert Competition.

Medical AICNNCheXpert
// Expertise

TechnicalDepth

Core competencies honed through years of building production AI systems at scale.

Large Language Models

Expert
RLHFDPOGRPOSFTSynthetic DataEvaluation
FK-GPT 8B outperformed GPT-4o-mini · 70B on parity with GPT-4o

Natural Language Processing

Expert
Machine TranslationSemantic SearchClassificationSummarization
Google & Azure Outperformed on translation

Conversational AI & Search

Expert
Search AgentsDialogueIntent DetectionVoice
+4% NDCG lift · 70% perfect slate results

Deep Learning & Vision

Advanced
CNNsGANsObject DetectionSegmentation
Top 30 in CheXpert Competition (Stanford)

ML Infrastructure

Advanced
EvaluationActive LearningReward ModelsData Gen
8M → 400M Scaled data parallel sentences

Product & Entrepreneurship

Expert
AI ProductsSocial GraphPersonalizationPrivacy
WizzMe Architected platform end-to-end
// Research

Publications

Peer-reviewed research at top-tier NLP venues.

ACL 2024
Main Conference

One Prompt To Rule Them All: LLMs for Opinion Summary Evaluation

Amey Patil, Tejpalsingh Siledar, Sudhanshu S. Singh, Muthusamy Chelliah, Nikesh Garera

Read Paper
EMNLP 2022
Industry Track

Large-scale Machine Translation for Indian Languages in E-commerce under Low Resource Constraints

Amey Patil, Nikesh Garera

Read Paper
NAACL 2024
Findings

Product Description and QA Assisted Self-Supervised Opinion Summarization

Amey Patil, Tejpalsingh Siledar, Banerjee, Sudhanshu S. Singh, M. Chelliah, N. Garera, Bhattacharyya

Read Paper
EMNLP 2023
Findings

Synthesize, if you do not have: Effective Synthetic Dataset Creation Strategies for Self-Supervised Opinion Summarization in E-commerce

Tejpalsingh Siledar, Suman Banerjee, Amey Patil, Sudhanshu Singh, M. Chelliah, N. Garera, Bhattacharyya

Read Paper
// Recognition

Awards

2024

Flipkart Value Champion

Building FK-GPT with 3x cost savings vs commercial LLMs

2023

Ace Alliance — Best DS Team

Exceptional contribution to Flippi chatbot and semantic search

2022

Flipkart Innovation Award

Machine translation platform for 11 Indian languages

2021

Flipkart Innovation Award

Translation platform expanding reach to India's linguistic communities

GitHub Arctic Code Vault

Code contributions archived for long-term preservation

2016

JEE Advanced AIR 65

All India Rank 65 — IIT Bombay Computer Science

// Education

AcademicFoundation

BTech Computer Science & Engineering

IIT Bombay

2016-2020

  • JEE Advanced AIR 65 · JEE Mains AIR 130
  • Research in AI for medical imaging
  • Core CS + AI/ML specialization

Intermediate / +2

Pace Jr. College Mumbai

2014-2016

  • Outstanding JEE performance
  • Foundation for IIT Bombay admission
// Open Source

GitHubProjects

github.com/ameypatil10

// Connect

Let's Build Something Together

Whether you want to discuss AI, explore collaboration, or just say hello.