Open to Product Management Opportunities & AI Consulting

Architecting Multi-Agent AI Systems That Scale

AI Systems Architect | Multi-Agent AI & Generative AI

Specialized in orchestrating 20+ agents, quality gating frameworks, and 0→50K user growth. Currently building JioCreate 3.0 at Jio Platforms - a 20-agent orchestration platform using Google Gemini 2.5 (Pro/Flash), Veo 3.0, and Imagen 4.0 for automated video production. Invented proprietary Audio Signature System and quality gating methodologies (8.1/10 threshold).

About Me

Akhil S - AI Product Manager

Turning AI Vision into Market Reality

With an MBA from IIFT Delhi and a B.Tech from NIT Agartala, I've spent the last 3+ years architecting AI products that matter. My journey began with building Avighna from scratch—a digital healthcare platform that grew from 2,000 to over 10,000 users under my product leadership.

Currently at Reliance, I'm pioneering AI initiatives across media and retail verticals, from content moderation systems to LLM integrations. My approach? Cut through the hype, focus on measurable impact, and build products that actually ship.

How I Build AI Products

1

Quality Gates Over Manual Review

I believe in automated quality systems, not armies of human reviewers. At JioCreate, I implemented 8.1/10 threshold across 6 dimensions with Critic-Refiner loops. Result: Production-ready output without manual QA bottlenecks.

2

Multi-Agent Systems Over Monolithic Models

Complex problems need specialized agents, not one GPT-4 call. I architect hierarchical systems (20 agents in JioCreate) where each agent masters one task. Result: Better quality + cost optimization vs. monolithic approaches.

3

0→1 with Small Teams, Scale with Systems

I start products solo or with 2-3 people (Avighna: solo → 3). Once product-market fit is proven, I build systems to scale (automated pipelines, frameworks, quality gates). Result: 25X growth without 25X team size.

4

Metrics That Matter Over Vanity Metrics

I optimize for retention, accuracy, and efficiency - not just user counts. Avighna: 55% bounce rate reduction matters more than 50K user headline. JioCreate: 8.1/10 quality threshold matters more than "20 agents" headline.

5

Build, Measure, Learn - But Learn Fast

I implement bounded iteration loops (max 2-3 refinements) to prevent analysis paralysis. Fast iteration with quality gates beats slow perfection. Result: 80% deployment success vs. ~50% industry average.

Looking for an AI Product Manager?

I'm currently exploring new opportunities in AI product management and available for consulting projects. Let's discuss how I can help drive your AI initiatives.

Professional Journey

2024 - Present
AI Product Manager
Reliance Industries Ltd.
Leading strategic AI initiatives across media and retail verticals. Developing AI censoring solutions, content recommendation systems, and LLM integrations for JioTV Plus and other platforms.
LLM Integration Media AI ROI Analysis
2022 - 2024
Product Manager - Avighna
Reliance Industries Ltd.
Built and scaled Avighna, a digital healthcare platform, from concept to 10,000+ users. Led integrations with JioHealthHub, MyJio, and ABDM ecosystem while achieving 60% accuracy in real-world medical trials.
Healthcare Tech 0 to 1 Product ABDM Integration
2018 - 2019
Developer
Primera Labs
Developed conversational learning platform reducing course duration by 40%. Enhanced UX by cutting navigation clicks from 7 to 3, improving user engagement significantly.
EdTech Frontend Dev UX Design
2016 - 2018
Senior Executive
Fiat India Automobiles
Led quality control initiatives for Tata Nexon project, achieving ₹15.46 Lakhs in cost savings. Reduced part rejection PPM from 1832 to 445 through process optimization.
Quality Control Process Optimization Kaizen

Featured Projects

Avighna Healthcare Platform
Digital Healthcare | AI Symptom Checker | 0→1 Product
Apr 2022 - Mar 2024 (2 years) | Team: Started solo → 3 people | Status: ✅ Live in Production
Built and scaled AI symptom checker from 0→50K users in 18 months (25X growth). Led product development solo for first 6 months, then scaled to team of 3 across web, mobile, WhatsApp deployment. Achieved 81% efficiency gain (content review 3 days→4 hours) and 17% diagnostic accuracy improvement in medical trials. Deployed across 3 platforms in 3 languages with integrations to JioHealthHub, Mylio (400M+ users), and ABDM ecosystem.
50K+
Users
25X
Growth
81%
Efficiency Gain
3
Medical Centers
AI Content Moderation System
Media AI | JioTV Plus
Aug - Dec 2023 (5 months) | Status: Completed PoC, pending production rollout
Developed AI-based censoring solution for JioTV Plus content compliance. Led project requirements definition, team formation, and end-to-end workflow design. Successfully demonstrated PoC with model meeting acceptance criteria for sensitive content detection. Achieved 81% efficiency improvement reducing content review time from 3 days to under 4 hours.
81%
Time Reduction
4 hrs
Review Time
PoC
Completed
LLM + Clinical Decision Support
Healthcare AI | Research
Led benchmarking of Avighna CDSS combined with Large Language Models for enhanced disease prediction. Managed data preparation, model evaluation, and performance reporting. Initial benchmarks showed improved accuracy with hybrid approach.
Improved
Disease Prediction
3
Model Variations
Benchmarked
Performance
Media AI ROI Framework
Strategic Planning | Portfolio Management
2024 - Present | Status: Framework enabled ₹150 Cr+ investment decisions with 80% deployment success
Established comprehensive ROI framework with financial modeling achieving <10% variance enabling ₹150 Cr+ investment decisions across AI portfolio. Evaluated 5+ solutions monthly with standardized prioritization methodology, increasing deployment success rate from ~50% baseline to 80%. Led ₹65 Cr AI portfolio with cross-functional team of 14, generating ₹110 Cr value monthly.
₹65 Cr
Portfolio Value
80%
Deployment Success
5+
Solutions/Month
14
Team Members
WhatsApp Healthcare Bot
Conversational AI | Healthcare
Designed and launched WhatsApp-based health assessment bot expanding Avighna's reach. Implemented multi-lingual support in Hindi and English. Created seamless conversational flows for symptom assessment via messaging.
3
Languages
24/7
Availability
Live
Status
State of AI Reports
Research | Thought Leadership
Created monthly AI intelligence reports analyzing trends across 7 industries. Provided strategic insights on emerging technologies and competitive landscape. Delivered actionable recommendations for Reliance's AI strategy.
3
Editions
7
Industries
50+
Insights

Trusted By

Jio Platforms
JioHealthHub
Mylio
ABDM

Recognition

Consecutive Star Performer - Reliance AI (FY 2023, 2024, 2025)

📊

₹65 Cr Portfolio Leadership - 80% deployment success rate

🎓

All India Top 1% CBSE Class XII | NSO AIR 698

🚀

15+ Production Features Shipped - Serving 450M+ users

Technical Expertise

Multi-Agent Orchestration

⚡ Expert

└─ JioCreate: 20-agent hierarchical system in production

Generative AI (LLMs, Video, Image)

⚡ Advanced

└─ Gemini 2.5 Pro/Flash, Veo 3.0, Imagen 4.0 deployed

0→1 Product Development

⚡ Expert

└─ Avighna: 0→50K users in 18 months (25X growth)

Quality Engineering

⚡ Advanced

└─ 8.1/10 threshold across 6 dimensions with automated gating

Healthcare AI

⚡ Expert

└─ 50K+ users, 3 medical center trials, 17% accuracy improvement

AI Systems Architecture

⚡ Expert

└─ Hierarchical agent design, quality gates, voice consistency frameworks

Portfolio Management

⚡ Expert

└─ ₹65 Cr portfolio, 80% deployment success, ₹150 Cr+ investment decisions

Cost Optimization

⚡ Advanced

└─ Multi-model strategy: 6 Gemini Pro + 14 Flash agents for cost-quality balance

Technology Stack

AI Platforms

Google ADK, Vertex AI (Gemini 2.5 Pro/Flash, Veo 3.0, Imagen 4.0), LangChain, LangFuse, FastMCP

Development

Python 3.12+, SQL, PostgreSQL, REST APIs, Docker, Kubernetes, Git, Postman

Analytics & Tools

GA4, Power BI, Azure DevOps, GCP (Cloud Storage, Agent Engine), Figma

Thoughts & Insights

Scaling Maslow: Tech Revolutions and Human Needs
Scaling Maslow: How Tech Revolutions Map to Our Deepest Human Needs
From survival calories to creative code, each innovation layer unlocks the next rung of human potential—and AI might just take us to the top and beyond.
Read on Substack →

My Journey

I've always been a systems thinker.

Started as a Quality Engineer at Fiat, optimizing manufacturing processes (saved ₹15.5L through Kaizen/defect reduction). Realized I wanted to build products, not just improve them.

Did my MBA at IIFT Delhi to learn business fundamentals, then joined Jio's AI Centre of Excellence in 2022 to apply systems thinking to AI.

Built my first 0→1 AI product: Avighna, a healthcare symptom checker. Started solo, scaled to 50K users in 18 months (25X growth), deployed across 3 platforms and 3 languages.

Now building my most complex system yet: JioCreate 3.0, a 20-agent AI platform for video generation. Invented the Audio Signature System, architected multi-model orchestration, and implemented quality gating frameworks that ensure production-ready output.

Leading a team of 14, managing ₹65 Cr AI portfolio with 80% deployment success.

What's next? I'm looking for principal PM or founding PM roles where I can architect novel AI systems at the frontier - whether that's AGI applications, multimodal AI, or the next paradigm shift.

Let's Build Something Amazing

Get In Touch

📧
Email
akhilrajsanthosh@gmail.com
💼
LinkedIn
linkedin.com/in/akhil-raj-b3531453
📱
Phone
+91 8732801743
📍
Location
Bangalore, India
🎯
Open to
Consulting | Speaking | Mentoring

Send a Message