Freelance AI Developers in Maharashtra

Freelance AI Developers in MaharashtraFreelance AI Developers in Maharashtra

Projects People

proMumbai, India

AI Video Producer | Brand Designer

$10k+: Earned

18x: Hired

5.0: Rating

249: Followers

expert

Expert

+4

AI Video Producer | Brand Designer

Cover image for AI Assistant Development for Sonic

AI Assistant Development for Sonic

2

23

Cover image for Support AI Agent for Cartelle

Support AI Agent for Cartelle

2

18

Cover image for The Foundry — AXIOM ONE

The Foundry — AXIOM ONE (15s spec spot) New spec drop: a cinematic launch film for a product I invented specifically to direct against — AXIOM ONE, a precision-milled titanium earbud. Process notes for anyone building in AI video: One 15s generation, but directed like a real spot: a locked global style prefix (cinematic / Kodak film emulation), a written product spec repeated verbatim, and three engineered cuts inside a single prompt. Continuity across the wider anthology comes from reference stills generated first, then fed into every video generation — the product never drifts. The "hyper-motion" feel is written as in-camera language: high-frame-rate capture, mixed shutter, engineered speed ramps — not "make it fast." This is Film II of four. Same product, four completely different stories, one visual DNA. Open for cinematic AI ad work — launch films, brand spots, product cinema.

1

30

Cover image for MAHABHARATA — FROZEN MOMENTS. The

MAHABHARATA — FROZEN MOMENTS. The epic's four turning points, suspended mid-instant in bronze: the dice throw, the unending sari, the charge at Kurukshetra, the bed of arrows. The reference format is frozen-time tableaux — 26 seconds, only 3 cuts, the world stopped at 2% speed while a single camera orbits through each moment with an ease-out. I reverse-engineered its grade before shooting anything: bronze-gold chiaroscuro, warm blacks, average frame brightness under 20% — one light source carving figures out of darkness. The build inverts normal production: • The keyframe is the film — each still sculpted in Nano Banana Pro until the frozen moment is perfect, suspension baked in (dice mid-air, shards mid-shatter, embers hanging) • Seedance 2.0 animates only the camera: one 360° orbit per 6-second shot, deceleration prompted in, landing on a named final frame • Four shots, soft dissolves, one continuous score — zero hard cuts • Reference screenshots fed as style conditioning so the cast-bronze material never drifts between shots Third entry in my Mahabharata series (after Suryaputra and The Bones). This is the pipeline I bring to launch and brand films — if you're building something that deserves mythology-grade imagery, Let's Talk!

4

262

AI Developer(2)

AI developer & backend engineer 💻

$1k+: Earned

1x: Hired

5.0: Rating

4: Followers

AI developer & backend engineer 💻

Cover image for ChatGPT Prompt Engineer

ChatGPT Prompt Engineer

0

46

Cover image for Building Conversational AI Chatbot

Building Conversational AI Chatbot

0

33

Cover image for Professional Technical Solution Consultant

Professional Technical Solution Consultant

0

14

Placeholder project card media

View more →

AI Developer(2)

I’m an AI & Machine Learning engineer with expertise in deve

I’m an AI & Machine Learning engineer with expertise in deve

Cover image for What your attention heatmap isn't

What your attention heatmap isn't telling you Everyone's staring at attention heatmaps and calling it "interpretability." Almost nobody's asking whether a single attention map actually tells you what the model used to make its decision. It doesn't. Not on its own. A raw attention map from layer 8 shows you what layer 8 attended to. It says nothing about how that signal got mixed, diluted, or overwritten by every layer before and after it. Attention rollout fixes this — and I built a walkthrough to show why it matters. Here's what makes it more than a "pretty heatmap" demo: Instead of visualizing one layer's attention, I traced how information actually flows through the full transformer stack. → Every layer's attention matrix is extracted, per head, per token → Multi-head attention is averaged, then combined with the residual connection (identity + attention) — this is the step most tutorials skip, and it's the one that actually matters → The combined matrices are matrix-multiplied layer by layer, rolling attention forward from input to output → The result: a single map showing genuine token-to-token influence across the entire network, not just one layer's snapshot The overlay shows you everything: → Per-layer attention vs. rolled-out attention, side by side → Token importance scores overlaid directly on the input text → A comparison view: which tokens raw attention says "matter" vs. which ones rollout says actually matter → Head-level breakdown so you can see which heads specialize vs. which are noise No black box. No "trust me, the model looked here." Just linear algebra, applied honestly across every layer instead of cherry-picking one. Built with PyTorch + HuggingFace Transformers + Matplotlib. Runs on any pretrained transformer, fully offline. ⚠️ Important: attention rollout is an approximation, not ground truth. It assumes attention is the primary information pathway, which ignores MLP layers and can still mislead for very deep models. Treat it as a debugging lens, not proof of causality.

0

30

Cover image for Everyone's racing to add biometrics

Everyone's racing to add biometrics to logins. Almost nobody's asking what happens when you can't — or shouldn't — touch the sensor. Shared kiosks, clinical settings, accessibility needs, hygiene-sensitive environments. Fingerprint readers and face unlock assume contact or a stored faceprint. Sometimes you want authentication that touches nothing and stores no biometric image of you at all. So I built GestureAuth — a contactless authentication system where your "password" is a sequence of hand gestures performed in front of a standard webcam.

1

28

Cover image for Everyone's talking about AI in

Everyone's talking about AI in healthcare. Nobody's building low-cost tools for the people who actually need early answers. Neurologists are overbooked. Clinical tremor assessments require in-person visits, specialist equipment, and months of waiting. The 10 million people living with Parkinson's globally and the millions more who don't yet know have no accessible way to flag early symptoms from home. So I built TremorLens a real-time hand tremor detection tool that runs entirely on a standard webcam. Here's what makes it more than just a webcam project: Instead of simple motion detection, I built a full signal processing pipeline on top of computer vision. MediaPipe tracks 21 hand landmarks per frame. The index fingertip's x/y displacement is buffered across a 3-second rolling window. scipy FFT then decomposes that signal into its frequency components and flags dominant activity in the 4–6 Hz range clinically associated with Parkinson's resting tremors. The live overlay shows you everything: → Real-time FFT power spectrum with the tremor zone highlighted → Dominant frequency readout in Hz with a 10-frame rolling average for stability → Color-coded STABLE / TREMOR DETECTED indicator → Fingertip displacement graph and movement trail → Auto-saved CSV session log timestamp, frequency, amplitude, tremor flag every session

0

25

Cover image for Everyone's talking about quantum computing.

Everyone's talking about quantum computing. Nobody's using it to feed farmers. India loses 20–30% of its crop yield every year to diseases and pests. Not because farmers don't care — but because early detection is hard, expensive, and inaccessible to the people who need it most. The existing solutions? Either a basic image classifier trained on lab-perfect photos that fail in real field conditions, or an agronomist visit that costs time and money most small farmers don't have. So I built QuantumEdge AgriGuard — a hybrid Quantum Neural Network app where a farmer can photograph a diseased leaf on their phone and get an instant diagnosis in under 5 seconds. Here's what makes it different from just another plant disease detector: Instead of a pure classical CNN, I built a hybrid architecture — a ResNet/EfficientNet backbone extracts visual features, then passes them into a Variational Quantum Circuit (VQC) for the final classification. The quantum layer uses angle embedding + StronglyEntanglingLayers, which gives it a measurable edge on small, noisy datasets — exactly the kind of data you get from Indian field conditions. The app doesn't just tell you what disease it is. It gives you: → Confidence score → Organic + chemical remedies (India-specific) → Yield impact estimate → A live classical vs quantum accuracy comparison so you can see the difference yourself I tested the quantum advantage claim honestly — ran both models on the same downsampled PlantVillage dataset and tracked accuracy, F1-score, and inference time side by side. The results are on the dashboard. No hand-waving. Built with PennyLane + PyTorch + Plotly Dash. Designed to run on simulators today and on QpiAI-Indus 25-qubit hardware tomorrow.

1

0

50

AI Developer(8)

Cloud Infrastructure Engineer | AWS, Serverless & AI Systems

Cloud Infrastructure Engineer | AWS, Serverless & AI Systems

Cover image for AI-Powered Document Summarizer & Chat System

AI-Powered Document Summarizer & Chat System

0

1

Cover image for Event-Driven Serverless Pipeline for Unstructured Data on AWS

Event-Driven Serverless Pipeline for Unstructured Data on AWS

0

3

Cover image for Clarity beats complexity.
I built this

Clarity beats complexity. I built this consequence map to cut through noise and see decisions for what they really are — trade-offs with outcomes. No guesswork, no overthinking. Just cause → effect, laid out clean. Good decisions aren’t luck. They’re mapped.

2

109

Cover image for PDFConverterOnlineFree – Node.js PDF Tool
I

PDFConverterOnlineFree – Node.js PDF Tool I built this project as a practical, real-world PDF tool that runs completely online and doesn’t rely on any database. The idea was simple — most PDF tools out there are either paid, slow, or store user files. I wanted to create something lightweight, fast, and privacy-friendly using open-source tools. 🔗 Project Link: https://github.com/deepakpatilauthor/PDF-Tool-Node-JS-Project What this project does This web app lets users: Convert PDFs to formats like Word, Excel, and images Convert files back into PDF Merge, split, and compress PDFs Use everything without signing up All files are processed temporarily and deleted automatically, so nothing is stored. How I built it I used Node.js for the backend and connected it with: Ghostscript for handling PDF operations like compression LibreOffice for document conversions The whole system is stateless (no database), which keeps it simple and fast. Why I built it this way I wanted to prove that you don’t always need paid APIs or heavy infrastructure to build something useful. With the right open-source tools, you can create a solid, production-ready app. What I learned Working with system-level tools like Ghostscript and LibreOffice Handling file uploads and processing efficiently in Node.js Building a clean workflow without relying on a database Thinking about performance and user privacy Final thoughts This project is a good example of how I approach building tools — keep it simple, make it useful, and avoid unnecessary complexity.

0

84

AI Developer(2)

Koustubh Lapate

Full Stack Developer

19: Followers

Full Stack Developer

Cover image for I am excited to announce that I have an Cattle Monitoring Sy...

I am excited to announce that I have an Cattle Monitoring System to detect lumpy disease in cows.!! Integrated real-time data collection on cow activity (steps count, temperature, milk production) via an API to assess health. Built a machine learning model for image classification to analyze cow images and predict lumpy disease. Implemented a continuous monitoring system, allowing farmers to upload images for daily health analysis over a period of time. Designed backend API to process data, perform AI-driven predictions, and send health alerts to the frontend.

1

173

Cover image for Hey everyone, Super excited to share something I’ve been wor...

Hey everyone, Super excited to share something I’ve been working on - inteGPT, an AI-powered platform I built to help interior designers and creative teams become 10x more productive. The goal behind inteGPT is simple: to help creatives focus more on imagination and less on repetitive work. The platform brings together speed, precision, and creativity, making it easier to generate ideas, designs, and assets in seconds. Whether you’re brainstorming design concepts or building client-ready visuals, inteGPT helps you stay ahead in today’s fast-moving creative world. Check it out and see how AI can completely transform your workflow - https://www.integpt.com/

1

266

Cover image for A website isn’t just your digital identity - it’s your first...

A website isn’t just your digital identity - it’s your first impression in milliseconds. I’ve seen clients focus entirely on visuals while ignoring usability. But a clean design without logical flow leads to drop-offs, not conversions. I recently redesigned a landing page to improve user journey clarity and load speed. The bounce rate dropped by 28%. My Portfolio - https://www.koustubhlapate.com/ Great design doesn’t just look good; it feels right to users. If you’re looking to refine your site’s UI/UX for real-world impact, I’m open for new freelance projects.

1

242

Cover image for I am excited to share that I’ve built an AI-based Real Estat...

I am excited to share that I’ve built an AI-based Real Estate Chatbot that helps users find properties, answer FAQs, and capture leads - all in real-time! 📌 Key Features: ✅ Instant responses to customer inquiries ✅ Smart property recommendations based on budget, location & other parameters ✅ Seamless lead capture to boost conversions 📽️ 𝐂𝐡𝐞𝐜𝐤 𝐨𝐮𝐭 𝐭𝐡𝐞 𝐝𝐞𝐦𝐨 𝐯𝐢𝐝𝐞𝐨 𝐚𝐭𝐭𝐚𝐜𝐡𝐞𝐝! 🚀 Want this chatbot for your real estate business? I’m ready to integrate it! 💡 Need a custom AI chatbot for your industry? I can build one tailored to your needs! Let’s connect and bring AI-powered automation to your business!

1

237

AI Developer(2)

Tech solutions with a product-first approach.

Tech solutions with a product-first approach.

Cover image for RewriterPro

0

33

Cover image for PhraseCraft: Revolutionize your web presence with AI-driven cont

PhraseCraft: Revolutionize your web presence with AI-driven cont

0

22

Cover image for MuchBeta: Connecting writers with tailored beta readers

MuchBeta: Connecting writers with tailored beta readers

0

13

Placeholder project card media

View more →

AI Developer(2)

AI Developer: Automate with AI Bots

AI Developer: Automate with AI Bots

Cover image for functionality of Perchance pages. This

functionality of Perchance pages. This

0

10

Cover image for functionality of Perchance pages. This

functionality of Perchance pages. This

0

3

Cover image for WIKI

0

10

Cover image for Text to image

0

10

AI Developer(4)

proMumbai, India

Framer Official Expert & Top 1% Creator

$1k+: Earned

3x: Hired

5.0: Rating

126: Followers

expert

Expert

Framer Official Expert & Top 1% Creator

Cover image for HealthTech CRM Application Prototype

HealthTech CRM Application Prototype

1

4

Cover image for Didn't overthink this hero design...

Clear

Didn't overthink this hero design... Clear message Classic visual direction Subtle motion that supports the story Built with @Unicorn Studio and a classic background that blends into the story through subtle shader effects.

2

3

345

Cover image for For our client, the goal

For our client, the goal was simple: create a better web experience. A familiar visual style @Unicorn Studio entry animations Thoughtful Mouseover micro-interactions -- A story that unfolds as you scroll Simple idea - Thoughtful execution.

2

3

312

Cover image for Where structured UI design meets intelligent motion.

Where structured UI design meets intelligent motion.

1

779

AI Developer(1)