Freelancers using LangChain in BengaluruFreelancers using LangChain in Bengaluru
ProjectsPeople
Vasundhhara Katoch's avatar
Vasundhhara Katoch
Bengaluru, India
AI Engineer delivering products from 0 to scale
New to Contra
AI Engineer delivering products from 0 to scale
Cover image for Personal Knowledge Model Development
Personal Knowledge Model Development
0
1
Cover image for AI Task Creation for Terminal-Bench
AI Task Creation for Terminal-Bench
0
0
Cover image for Khoros Iris Web Application Development
Khoros Iris Web Application Development
0
0
Cover image for Enhancements to Maestro for Parallel Software Sessions
Enhancements to Maestro for Parallel Software Sessions
0
0
LangChain(1)
Mann Acharya's avatar
Mann Acharya
proBengaluru, India
AI Product Engineer: Ideation, Development, Deployment
New to Contra
AI Product Engineer: Ideation, Development, Deployment
Cover image for “All AI agents are just
“All AI agents are just the same 3 steps.” Reason → Act → Observe. (ReAct — Yao et al., ICLR 2023) (https://arxiv.org/pdf/2210.03629)The challenge is not development. It’s knowing that your system actually works, reliably, in production. How does it handle edge cases? Is it safe? Does it stay on task? How repeatable are the outputs? What’s the delay? What’s the price? Strong evaluations are necessary here. Evals turn impressive demos into battle tested AI products that teams can really count on and scale. Working on AI products where reliability actually matters? Let’s connect. (https://contra.com/mann_acharya/work)– Mann, AI Product Engineer: Ideation, Development, Deployment
2
2
172
Cover image for Social Media Intelligence & Monitoring Platform
Social Media Intelligence & Monitoring Platform
0
5
Cover image for Enterprise Code AI Assistant
Enterprise Code AI Assistant
0
10
Cover image for Health Application
Health Application
0
8
LangChain(3)
Sowmya Lanka's avatar
Sowmya Lanka
Bengaluru, India
Data Scientist | ML, NLP & GenAI Solutions
New to Contra
Data Scientist | ML, NLP & GenAI Solutions
Cover image for AI-Powered Policy Q&A Assistant Built a
AI-Powered Policy Q&A Assistant Built a Retrieval-Augmented Generation (RAG) chatbot using Azure OpenAI GPT to answer policy-related questions with accurate, context-aware responses. What I built Extracted and structured information from policy documents into a knowledge base. Used Azure Cosmos DB for scalable information storage and retrieval. Implemented RAG to retrieve relevant policy information based on user queries. Integrated Azure OpenAI GPT to generate natural-language, context-aware answers. Technologies Python · Azure OpenAI · GPT · RAG · Azure Cosmos DB · LangChain Outcome The solution enables users to quickly find relevant information from policy documents without manually searching through lengthy documents.
1
92
Cover image for What I built Developed a machine
What I built Developed a machine learning solution to predict Buy Box and sales prices using historical analytical data and Linear Regression. How it works Analyzed the data for linear relationships, autocorrelation, multicollinearity, homoscedasticity, and normally distributed errors. Applied Principal Component Analysis (PCA) for dimensionality reduction and to improve model performance. Identified key business factors influencing price prediction. Evaluated the model using R², Adjusted R², RMSE, MAE, and MSE metrics. Technologies Python · Pandas · NumPy · Scikit-learn · Linear Regression · PCA Outcome Built a price prediction model that identified important business drivers and evaluated prediction performance using multiple regression metrics.
1
84
Cover image for What I built Developed a Natural
What I built Developed a Natural Language Processing (NLP) solution to detect DI Flag issues by analyzing and classifying text descriptions. How it works Cleaned text data by removing hashtags, HTML tags, special characters, and numeric values. Applied text preprocessing techniques including tokenization, stop-word removal, stemming, and lemmatization. Converted text into numerical features using Bag of Words (BoW) and TF-IDF. Trained and evaluated the model using accuracy and confusion matrix metrics. Performed 30 days of validation testing before production deployment. Technologies Python · NLP · Scikit-learn · TF-IDF · Bag of Words · Pandas · NumPy Outcome Built a text classification solution that helped identify DI Flag issues from descriptions and validated the model's performance before deployment.
1
82
Cover image for What I built Developed an automated
What I built Developed an automated clinical text summarization solution using Google Gemini AI and prompt engineering to generate concise summaries from lengthy clinical chart notes, transcripts, and unstructured medical text. How it works Processed lengthy clinical notes and transcripts. Used prompt engineering to guide Gemini in identifying important medical information. Extracted key details such as symptoms, diagnoses, medications, and treatment plans. Generated concise, readable summaries from the original clinical text. Technologies Google Gemini AI · Python · Prompt Engineering · NLP · Generative AI Outcome Reduced clinical chart review time by approximately 30%, helping healthcare professionals identify important information more quickly and efficiently.::
1
74
LangChain(1)
Vimal Anand's avatar
Vimal Anand
Bengaluru, India
Full Stack AI Engineer shipping fast SaaS MVPs
New to Contra
Full Stack AI Engineer shipping fast SaaS MVPs
Cover image for Ever watched an AI write
Ever watched an AI write bad SQL, instantly realize its mistake, and rewrite it perfectly—all without human intervention? Over the weekend, I built the Autonomous Data Analyst Agent. It doesn’t just translate natural language to SQL; it actively debugs itself. 𝗧𝗵𝗲 𝗣𝗿𝗼𝗯𝗹𝗲𝗺: Enterprises want AI to query their proprietary databases, but LLMs often hallucinate columns or mess up syntax. 𝗧𝗵𝗲 𝗦𝗼𝗹𝘂𝘁𝗶𝗼𝗻: I engineered a Self-Correcting Reflection Loop using LangGraph. Here’s what happens under the hood: 1️⃣ You ask a complex question in plain English. 2️⃣ The LLM generates DuckDB-flavored SQL. 3️⃣ The Executor node runs it securely in a local container. 4️⃣ 🔄 𝗧𝗵𝗲 𝗠𝗮𝗴𝗶𝗰: If DuckDB throws a parser error, the Reflection node catches it, feeds the error trace back to the LLM, and auto-corrects the query. 5️⃣ Once successful, a Python agent generates a Matplotlib visualization, streaming everything to a sleek Next.js dark-mode UI. Plus, I integrated LangGraph MemorySaver, allowing the agent to retain context for human-like follow-up questions. 𝗧𝗲𝗰𝗵 𝗦𝘁𝗮𝗰𝗸: Next.js, FastAPI, LangGraph, DuckDB, Groq (Llama-3), Docker Compose. Everything is fully containerized and open-source. Drop a ⭐ on the repo and let me know what you think! 𝗚𝗶𝘁𝗛𝘂𝗯 𝗥𝗲𝗽𝗼: https://lnkd.in/ddnE9rgq (https://lnkd.in/ddnE9rgq)hashtag#AI (https://www.linkedin.com/search/results/all/?keywords=%23ai&origin=HASH_TAG_FROM_FEED) hashtag#LangGraph (https://www.linkedin.com/search/results/all/?keywords=%23langgraph&origin=HASH_TAG_FROM_FEED) hashtag#DataEngineering (https://www.linkedin.com/search/results/all/?keywords=%23dataengineering&origin=HASH_TAG_FROM_FEED) hashtag#Nextjs (https://www.linkedin.com/search/results/all/?keywords=%23nextjs&origin=HASH_TAG_FROM_FEED) hashtag#FastAPI (https://www.linkedin.com/search/results/all/?keywords=%23fastapi&origin=HASH_TAG_FROM_FEED) hashtag#Docker (https://www.linkedin.com/search/results/all/?keywords=%23docker&origin=HASH_TAG_FROM_FEED) hashtag#MachineLearning (https://www.linkedin.com/search/results/all/?keywords=%23machinelearning&origin=HASH_TAG_FROM_FEED) hashtag#OpenSource (https://www.linkedin.com/search/results/all/?keywords=%23opensource&origin=HASH_TAG_FROM_FEED)
1
30
Cover image for 𝗛𝗲𝗮𝗱𝗹𝗶𝗻𝗲: What started as a
𝗛𝗲𝗮𝗱𝗹𝗶𝗻𝗲: What started as a simple B2C expense tracker has officially pivoted into a fully automated B2B Financial SaaS. Introducing 𝗙𝗶𝗻𝗮𝗻𝗦𝗺𝗮𝗿𝘁. 𝗕𝗼𝗱𝘆: I realized that startup founders and agency owners spend countless hours manually categorizing bank statements and tracking operational costs. I wanted to build a solution that entirely automates this bookkeeping process. Building the core feature—an "𝐀𝐈 𝐁𝐚𝐧𝐤 𝐒𝐭𝐚𝐭𝐞𝐦𝐞𝐧𝐭 𝐏𝐚𝐫𝐬𝐞𝐫"—was one of the toughest technical challenges I’ve faced. I hit a massive roadblock when Next.js 14’s Webpack bundler kept breaking legacy PDF OCR libraries in serverless environments. Instead of compromising on the feature, I completely re-architected the data pipeline: • I bypassed Webpack issues by implementing pdf2json on a strict Node.js runtime for secure, server-side text extraction. • I piped this raw text into Groq’s LLaMA-3.1-8b model using highly optimized system prompts. • The result? Lightning-fast, deterministic extraction that converts raw PDF text into perfectly structured JSON arrays. Now, users can drag-and-drop a PDF statement, and the AI instantly categorizes every transaction (Infrastructure, Payroll, Revenue) into a 'pending' staging area. Once approved, the data flows securely via Drizzle ORM into a Neon PostgreSQL database, updating real-time Recharts dashboards. 𝑻𝒆𝒄𝒉 𝑺𝒕𝒂𝒄𝒌: Next.js App Router, Tailwind CSS, Shadcn UI, Clerk Auth, Drizzle ORM, Neon DB, and Groq SDK. 𝗪𝗵𝗮𝘁'𝘀 𝗡𝗲𝘅𝘁? I am currently looking for my next full-time opportunity as a Full Stack Developer (based in Bangalore or anywhere). I am also actively taking on 𝗳𝗿𝗲𝗲𝗹𝗮𝗻𝗰𝗲 𝗽𝗿𝗼𝗷𝗲𝗰𝘁𝘀. If you are a founder or agency looking to build scalable, AI-integrated SaaS MVPs without the technical headache, let's connect. Check out the demo video below to see the AI parser in action! 🔗 Live Project: https://lnkd.in/gSNu3cAK (https://lnkd.in/gSNu3cAK)hashtag#BuildInPublic (https://www.linkedin.com/search/results/all/?keywords=%23buildinpublic&origin=HASH_TAG_FROM_FEED) hashtag#Nextjs (https://www.linkedin.com/search/results/all/?keywords=%23nextjs&origin=HASH_TAG_FROM_FEED) hashtag#FullStackDevelopment (https://www.linkedin.com/search/results/all/?keywords=%23fullstackdevelopment&origin=HASH_TAG_FROM_FEED) hashtag#SaaS (https://www.linkedin.com/search/results/all/?keywords=%23saas&origin=HASH_TAG_FROM_FEED) hashtag#AI (https://www.linkedin.com/search/results/all/?keywords=%23ai&origin=HASH_TAG_FROM_FEED) hashtag#Groq (https://www.linkedin.com/search/results/all/?keywords=%23groq&origin=HASH_TAG_FROM_FEED) hashtag#PostgreSQL (https://www.linkedin.com/search/results/all/?keywords=%23postgresql&origin=HASH_TAG_FROM_FEED) hashtag#WebDevelopment (https://www.linkedin.com/search/results/all/?keywords=%23webdevelopment&origin=HASH_TAG_FROM_FEED) hashtag#FreelanceDeveloper (https://www.linkedin.com/search/results/all/?keywords=%23freelancedeveloper&origin=HASH_TAG_FROM_FEED) hashtag#Hiring (https://www.linkedin.com/search/results/all/?keywords=%23hiring&origin=HASH_TAG_FROM_FEED)
1
39
Cover image for I recently architected and built
I recently architected and built VedaAI Assessment Creator, a personal project designed to transform unstructured notes and files (PDFs, DOCX) into highly structured, curriculum-aligned exam papers. Building enterprise-grade applications at scale requires looking beyond simply making API calls. For this build, I wanted to focus entirely on non-blocking architectures and deterministic AI outputs. Here is a breakdown of the technical decisions: Asynchronous Workers: Instead of blocking the main HTTP thread during heavy file parsing and AI inference, I implemented a distributed task queue using BullMQ and Upstash Serverless Redis. The Express API responds in under 100ms, while background workers handle the heavy lifting. Deterministic AI: I opted for Groq (Llama-3.3-70b) utilizing its JSON mode. The sub-500ms inference time and guaranteed schema compliance eliminated the need for complex post-processing validation. Real-Time Synchronization: A WebSocket setup broadcasts job completion events, updating the Next.js frontend instantly without relying on inefficient polling. Database & State: MongoDB Atlas handles ACID transactions for complex document updates, while Zustand manages lightweight, atomic state slices on the frontend. I also built in granular question regeneration—allowing users to re-run isolated inference calls for single questions without replacing the entire paper—and print-ready A4 PDF exports. Designing this kind of scalable infrastructure directly supports my ongoing deep dive into advanced AI and machine learning systems. You can check out the Live Link here: https://lnkd.in/geAXUtHC . I would love to hear how others are handling asynchronous AI tasks in production!
1
25
Cover image for 𝗛𝗲𝗮𝗱𝗹𝗶𝗻𝗲: Most Voice AI agents
𝗛𝗲𝗮𝗱𝗹𝗶𝗻𝗲: Most Voice AI agents still feel like glorified walkie-talkies. So I built one that actually listens like a human. Most voice bots today follow a rigid, linear loop: You speak ➔ Wait ➔ Bot speaks. If you dare interrupt them mid-sentence? They either completely ignore you or their context buffer gets totally mangled. 𝗧𝗼 𝘀𝗼𝗹𝘃𝗲 𝘁𝗵𝗶𝘀, I engineered Astra Duplex Agent — a sub-50ms, ultra-low latency, full-duplex conversational Voice AI built to handle natural human interruptions seamlessly. Here is what went into building the architecture under the hood: True Full-Duplex WebSockets: Built on FastAPI, allowing real-time bi-directional audio streaming instead of turn-based HTTP requests. Edge Voice Activity Detection (VAD): Deployed Silero VAD using ONNX Runtime to catch user barge-ins at the edge with near-zero overhead. LangGraph State Machine: Instead of a simple monolithic script, the conversation logic is modeled as a state graph, giving precise control over execution flow. Partial State Tracking & Memory Persistence: This was the trickiest part. When an interruption happens, Astra doesn't just stop audio playback — an async kill-switch halts Groq's LLaMA-3.1 mid-token, calculates the exact partial sentence actually spoken by the TTS, and saves only that partial context to an Upstash Redis buffer. Sub-50ms Inference & Streaming: Leveraged Groq LPU (LLaMA-3.1 + Whisper Large v3) paired with ElevenLabs Turbo v2.5 streaming. The result? An agent that you can interrupt mid-thought, change topics with on the fly, and ask follow-up questions without it losing historical context. Containerized with Docker and live on Render & Vercel. Check out the full architecture & live demo below! https://lnkd.in/eC6yak3i (https://lnkd.in/eC6yak3i)https://lnkd.in/eN38HGZD (https://lnkd.in/eN38HGZD)#VoiceAI (https://www.linkedin.com/search/results/all/?keywords=%23voiceai&origin=HASH_TAG_FROM_FEED) #GenerativeAI (https://www.linkedin.com/search/results/all/?keywords=%23generativeai&origin=HASH_TAG_FROM_FEED) #LangGraph (https://www.linkedin.com/search/results/all/?keywords=%23langgraph&origin=HASH_TAG_FROM_FEED) #SystemDesign (https://www.linkedin.com/search/results/all/?keywords=%23systemdesign&origin=HASH_TAG_FROM_FEED) #FastAPI (https://www.linkedin.com/search/results/all/?keywords=%23fastapi&origin=HASH_TAG_FROM_FEED) #MachineLearning (https://www.linkedin.com/search/results/all/?keywords=%23machinelearning&origin=HASH_TAG_FROM_FEED) #WebSockets (https://www.linkedin.com/search/results/all/?keywords=%23websockets&origin=HASH_TAG_FROM_FEED) #Python (https://www.linkedin.com/search/results/all/?keywords=%23python&origin=HASH_TAG_FROM_FEED) #AIENGINEERING (https://www.linkedin.com/search/results/all/?keywords=%23aiengineering&origin=HASH_TAG_FROM_FEED)
1
58
LangChain(1)
Azeem Husain Khan's avatar
Azeem Husain Khan
Bengaluru, India
AI Engineer | RAG & Agentic Systems | RAG | Chatbots
AI Engineer | RAG & Agentic Systems | RAG | Chatbots
Cover image for A satirical web app that
A satirical web app that generates a shareable "Certificate of Heroic Weekend Sacrifice" — parody of protest-attendance certificates, built entirely around a fictional movement.
0
6
Cover image for AI Avatar Generator is a
AI Avatar Generator is a Generative AI application that allows users to create personalized AI avatars from text prompts. The project combines image-generation models with an interactive web interface, enabling users to customize avatar styles, appearance, and creative concepts. I designed, integrated, and deployed the application with a focus on prompt engineering, AI model integration, and an intuitive user experience.
0
13
Cover image for Built an intelligent LLM routing
Built an intelligent LLM routing system that dynamically discovers and evaluates available models, performs health and rate-limit checks, and routes each request to the most suitable model based on capability, latency, availability, and cost. The system supports automatic failover and manual model selection through a Streamlit dashboard, providing a practical AI gateway for reliable and cost-efficient LLM applications.
0
15
Cover image for EasyPeasyTales.com
EasyPeasyTales.com
0
40
LangChain(3)
Jyotirmaya Sahu's avatar
Jyotirmaya Sahu
Bengaluru, India
Frontend architect with good knowledge in backend and AI
Frontend architect with good knowledge in backend and AI
Cover image for Hi there 👋 I’m a Full-Stack / Frontend Engineer with strong...
Hi there 👋 I’m a Full-Stack / Frontend Engineer with strong experience in React, dashboards, performance optimization, and applied AI. I’ve shared my portfolio and resume here — happy to discuss how I can help with your project. Looking forward to connecting!
0
15
Cover image for Hey everyone! I’m excited to share something I’ve been build...
Hey everyone! I’m excited to share something I’ve been building behind the scenes — a fully custom AI-driven SaaS dashboard that helps businesses turn raw data into clear insights, real-time metrics, and actionable decisions. Here’s the live demo 👇 🔗 https://saas-dashboard-lime.vercel.app/login (https://saas-dashboard-lime.vercel.app/login)Demo login is available on the site.
0
40
Cover image for Hey folks! 👋 I’m planning to build an AI-powered proposal g...
Hey folks! 👋 I’m planning to build an AI-powered proposal generator — but not the generic ones that exist today. This one will: ✨ Learn from your portfolio ✨ Study which proposals worked for you ✨ Get smarter with every project you complete ✨ Generate highly personalized, high-conversion proposals Basically: your own proposal-writing co-pilot. Would love feedback or feature ideas!
0
19
Cover image for I am building this side project to simplify my long podcast ...
I am building this side project to simplify my long podcast list https://www.pulseread.app/
0
22
LangChain(2)