Skip to main content
TJTaaran Jain

OpenOpen to AI Engineer roles · Remote or on-site

AI Engineer · LLMs & Applied ML · Full-StackTaaran Jain

I build RAG and agentic AI apps, from model to screen.

I'm an AI Engineer with two years of hands-on experience across 3 companies and 10+ industries. I take AI products from first idea to launch, and I build them to hold up after the demo, not just during it.

See Live Projects View Resume

2+

Years of professional ML & DL work

7

Live AI products to try

2

Research papers published

Illustrative trace of how Nexus AI answers: route, retrieve, generate, cite.Try it live

Experience2+ years · 3 companies · 10+ industries

  • Python
  • LangGraph
  • LangChain
  • RAG
  • PyTorch
  • scikit-learn
  • FastAPI
  • Next.js
  • TypeScript
  • pgvector
  • Hugging Face
  • Groq
  • Docker

01About

An AI Engineer Who Builds the Whole App

Give me a problem and I'll take it all the way to a product people can use. 7 of them are live on this site.

I'm Taaran, an AI Engineer in Jaipur, India. I started in data science, moved into machine learning and deep learning on production systems, and now build LLM products: retrieval systems that cite their sources and agents that decide which tool to use.

The problems I enjoy most are the ones where a model has to earn trust: an answer that needs a source, a prediction someone will act on, an alert that can't cry wolf. That's why I build the backend and the interface myself, so I can see how the AI behaves for the person actually using it.

I make trade-offs deliberately. A smaller model if it's fast enough, files on disk when one server will do, a rule-based fallback when the AI is slow. Each case study shows those choices, the numbers behind them and what I'd change next time.

PythonLangGraphLangChainRAGpgvectorHugging FacePyTorchscikit-learnFastAPINext.jsTypeScriptDockerGCP

LLM & RAG Engineering

RAG pipelines that cite the exact page and line, LangGraph agents that pick which tool to call, and one app that runs on 8 LLM providers.

Machine & Deep Learning

ML and deep learning anomaly detection on production logs, neural networks that learn to drive, calibrated churn models, and a workbench comparing 68 algorithms.

Full-Stack Engineering

FastAPI backends, Next.js and TypeScript front ends, Docker deploys, and the rate limits and quotas that let strangers use a free demo safely.

Numbers Over Adjectives

Held-out evaluation, calibration error, and latency and token cost per request. The case studies also cover what failed the first time.

7

AI products live with demos

3

Companies shipped for

10+

Industries served

2

Research papers published

02Projects

AI Projects You Can Try Live

Each one solves a real problem and is live to try: an AI assistant that cites its sources, a churn model that tells teams who's about to leave, a log analyser that caught 588 login attacks, and 4 more.

nexus-ai.taaranjain.comLive
01RAG · LangGraph Agent

Nexus AI

An AI assistant that shows its sources and confidence

Problem
Chat assistants give you an answer and hide the rest: whether it came from the web, your files or the model's memory, and how confident it is. That's exactly what matters when you're checking a policy or making a real decision.
What I Built
A six-node LangGraph pipeline routes each question to web search, your documents or the model, and streams every step with its time and token cost. Answers cite the exact file, page and line, and a calibrated decision model scores every option when you're choosing.
Result
Every answer is auditable: route, sources, timing and token cost stream live alongside the reply

0.4s

Decision board scored

~2×

Faster cold start

8

LLM providers

How it was measured: Local testing with real API keys. Cold start is the backend's import time before and after the change.

  • LangGraph
  • LangChain
  • Groq
  • FastAPI
  • Next.js
  • TypeScript
  • Hugging Face
Try the Live Demo Read the Case Study
decidex.taaranjain.comLive
02ML Churn Prediction

DecideX

Tells customer teams who might leave, why, and what to do

Problem
Customer teams usually find out a customer is leaving when it's too late, and a churn score on its own doesn't say what to do next, who should do it, or how much the account is worth saving.
What I Built
A decision app that reads account notes with an LLM and scores each customer with a calibrated churn model, then recommends what to do, who owns it and how urgently, with plain-English reasons, the revenue at risk and the smallest offer that brings the customer back under the risk line.
Result
Catches 85.6% of leavers on 1,409 held-out customers from the IBM Telco dataset (ROC-AUC 0.844, PR-AUC 0.653), with probabilities calibrated to an ECE of 0.018

85.6%

Leavers caught on held-out data

−71%

Missed churners after cost tuning

0.018

Calibration error (ECE)

How it was measured: Public IBM Telco dataset (7,043 customers), stratified 60/20/20 split, scored on the 1,409 held-out customers.

  • Python
  • scikit-learn
  • LightGBM
  • FastAPI
  • Next.js
  • TypeScript
  • SQLite
Try the Live Demo Read the Case Study
logvera.taaranjain.comLive
03ML Anomaly Detection

Logvera

Turns raw logs into incidents, attacks and root causes

Problem
Reading a raw log to find out what went wrong is slow, and most log tools need agents, accounts and setup before they show you anything. Engineers end up grepping by hand, and an attack spread across thousands of lines is easy to miss.
What I Built
Drop in a log or an archive and Logvera detects the format, learns message patterns with a Drain-style parser, and runs a five-model anomaly ensemble plus a rule-based security scan. An LLM layer writes summaries and answers questions with cited lines, seeing only redacted data, and every upload is encrypted and deleted after an hour.
Result
On a 34,492-line sample log it found 3 unusual periods, 8 security findings including 588 credential-stuffing attempts, and 84 distinct message patterns

5

Models voting on anomalies

588

Attacks found in sample log

41

Backend tests

How it was measured: A bundled sample of one day of e-commerce Nginx traffic with planted incidents, not a production log.

  • Python
  • FastAPI
  • scikit-learn
  • NumPy
  • Next.js
  • TypeScript
  • WebSockets
Try the Live Demo Read the Case Study

More Shipped Projects

All Case Studies

03Experience

Two Years Shipping AI in Production

Data Science at Celebal Technologies, Anomaly Detection at LogiScope and AI Products at e-Marketing.io, with client work across 10+ industries.

2+

Years of professional ML & DL work

3

Companies

From data science to deep learning on production systems to LLM products, with client work across 10+ industries along the way.

  1. AI Engineer

    e-Marketing.io

    CurrentJun 2025 – Present

    Design and ship the AI products a performance marketing agency runs for its clients, owning each one from problem scoping and model selection to deployment and iteration.

    • Built an AI meeting summarizer that turns hour-long recordings into structured briefs with decisions and action items, cutting follow-up work after every call.
    • Shipped lead-qualifying AI chatbots across social and messaging platforms that reply instantly and keep client pipelines active around the clock.
    • Developed an NLP keyword analysis engine that surfaces trends, search intent and competitor gaps, replacing gut-feel content planning.
    • Delivered a lead management dashboard with real-time pipeline status and full bot conversation history for each client.
    • Built a WhatsApp task delegation system where teams assign, track and close tasks by text or voice note, without switching apps.
    • Python
    • LangChain
    • OpenAI API
    • WhatsApp Business API
    • FastAPI
    • React
    • PostgreSQL
    • Docker

  2. Data Science Specialist

    LogiScope Technologies Pvt. Ltd.

    Dec 2024 – Jun 2025

    Built ML and deep learning backends that turned large, noisy system logs into alerts that monitoring and reliability teams could act on.

    • Built and deployed anomaly detection models, using both ML and deep learning, that flag system irregularities in real time before they escalate.
    • Ran EDA on large-scale log datasets and surfaced failure patterns that manual monitoring consistently missed.
    • Benchmarked several detection techniques to find the most reliable signals in complex log behaviour.
    • Integrated the findings into data processing pipelines and improved the anomaly reporting framework with cross-functional teams.
    • Python
    • PyTorch
    • Scikit-learn
    • Pandas
    • NumPy
    • Deep Learning
    • Anomaly Detection
    • EDA

  3. Data Science Intern

    Celebal Technologies

    May 2024 – Aug 2024

    Worked in a production data science team, taking projects from raw data to deployed models on Azure.

    • Cleaned and preprocessed large datasets so the models trained on reliable inputs.
    • Implemented ML models against real business problems, from first experiment to a working predictive model.
    • Used Azure cloud services to see how production data applications are architected and deployed.
    • Delivered cross-functional project work alongside senior data scientists.
    • Python
    • Pandas
    • NumPy
    • Matplotlib
    • Scikit-learn
    • Azure
    • Machine Learning

04Research

Published Research

Papers on applied deep learning: reinforcement learning for trading, and multi-modal AI for medical diagnosis.

05Skills

Skills & Tech Stack

Each skill shows where I've used it, in a project or at work, and how much: every day, in production, or just familiar.

How I build an LLM application

  1. 01 · LanguagePython
  2. 02 · OrchestrationLangGraph
  3. 03 · RetrievalRAG · pgvector
  4. 04 · ModelsPyTorch · sklearn
  5. 05 · APIFastAPI
  6. 06 · InterfaceNext.js
  7. 07 · ShipDocker · GCP
Daily=Daily driverProduction=Shipped in productionFamiliar=Working knowledge

LLM / NLP

8

Machine Learning

5

Deep Learning

3
  • PyTorchProduction

    Used in LogiScope

  • TensorFlow / KerasProduction
  • OpenCVProduction

Languages

2

Backend & APIs

4

Databases

4

Cloud & DevOps

3

Data Visualisation

3
  • Matplotlib / SeabornDaily

    Used in Celebal

  • PlotlyProduction
  • Power BIFamiliar

06Leadership

Leadership & Community

10 roles across GDSC, AICTE, HackerEarth and the student council, from leading the AI/ML team at GDSC to running hackathons and mentoring students.

07Certifications

Certifications

A WorldQuant Challenge silver, plus 10 credentials in machine learning, deep learning, MLOps and analytics from DeepLearning.AI, Google, Microsoft and ISB.

08Contact

Hiring an AI Engineer? Let's Talk.

I'm looking for full-time AI Engineer, Machine Learning Engineer, Full Stack Engineer or Data Scientist roles. Remote or on-site both work, and I'm happy to relocate for the right team.

  1. 01Send the Role or the ProblemA job description, a link or a few lines is enough.
  2. 02A Reply Within a DayI confirm fit and availability, and answer anything you need from me.
  3. 03Intro or Technical CallHappy to walk through any project on this site in depth, code included.
  • Replies usually within a day
  • Jaipur, Rajasthan, India · IST (UTC+5:30)

Goes straight to my inbox and is never stored on this site. I usually reply within a day. Privacy

Or reach me directly

FAQ

Questions Recruiters Ask

The things recruiters usually ask me first: roles, location, stack and experience.

Still have a question?

Ask directly. I usually reply within a day.

Ask a Question Email

Full-time AI Engineer, Machine Learning Engineer, Full Stack Engineer or Data Scientist roles. I'm also open to freelance projects and research collaborations if the problem is interesting.