Open to Data Science, Applied AI & Data Engineering roles
Data Science, Applied AI & Data Engineering—built for production.
Curious by nature and ambitious about building products, I like taking APIs, data flows, and models apart to understand why they work—and where they fail. I bring that mindset to airline data platforms, pricing models, machine learning, and observable, governed AI systems with evaluation, human oversight, and production reliability.
1.5+ yrs
Enterprise Experience
MT
Mohit Taneja
Data Science · Applied AI · Data Engineering
Experience & Independent Work
Two tracks, one trajectory.
My professional track is grounded in enterprise data and AI systems. In parallel, my independent track has pushed me to build products end to end, work directly with clients, and develop the judgment needed to ship AI responsibly.
Professional trajectory
Enterprise roles
AUG 2024 – JAN 2026 · GURUGRAM
Data Engineer
IndiGo (InterGlobe Aviation Ltd)
- Owned the Flight Operations migration from SQL Server to Snowflake as the sole data engineering point of contact, translating business requirements into procedures, functions, pipelines, and Dynamic Tables and delivering on schedule.
- Used Python, PySpark, and SQL for distributed processing; built Snowflake and Azure Data Factory pipelines across enterprise systems and data lakes, reducing processing overhead by 40%.
- Integrated Flight Operations, AIMS, crew/roster, and Navitaire sales data for flight-delay analysis, prediction, forecasting, and downstream data science; validated datasets and investigated source-to-target quality issues.
- Developed pricing and AI-assisted database administration POCs, testing models on Navitaire sales data for ticket and ancillary-bundle pricing, including XL seats and priority services.
- Co-built AskPrism, a RAG assistant connecting five business systems and used by 200+ employees, reducing resolution time by roughly 60% and manual analyst effort by roughly 45%.
- Administered Snowflake grants, revocations, RBAC, and PII access; supported database observability, controlled access, and production incident troubleshooting.
- Applied stored procedures, Dynamic Tables, materialized views, Streams, Time Travel, and Jenkins deployments; delivered Power BI dashboards and contributed to Tableau visualizations.
FEB 2024 – JUL 2024 · GURUGRAM
AI Researcher (Internship)
Siemens Energy
- Benchmarked enterprise LLM and NLP options across accuracy, cost, and operational complexity and authored the recommendation adopted by senior engineering leadership.
- Built a production assistant with Azure OpenAI and Assistant APIs that automated 3+ engineering workflows at 89% intent-classification accuracy.
- Used Azure and Copilot to integrate new chatbot capabilities and accelerate development and workflow automation.
Independent trajectory
Products & clients
FEB 2026 – PRESENT · REMOTE
Independent AI/ML Engineer
Client & Product Work
- Build AI-powered products and workflow automation systems using LLMs, APIs, tool calling, multi-agent orchestration, and human-in-the-loop architecture.
- Developed an AI-assisted Adobe Premiere Pro system combining model reasoning with deterministic automation, external processing services, and editor-controlled execution.
- Designed specialized-agent workflows with persistent context, tool access, state management, fallback paths, structured outputs, and reusable decision memory.
- Built integrations across REST APIs, MCP servers, OpenAI, Gemini, Claude, Codex, media-processing services, and creative tools using Python and TypeScript.
- Designed evaluation and control loops with structured tests, traces, human approval, fallback behavior, and multi-model review to improve AI observability, governance, and reliability.
- Worked through real production constraints including hallucinations, context loss, API limits, model cost, latency, unreliable tool execution, and nondeterministic output.
2025 · REMOTE
AI Developer (Contract)
Tech Baton · Agiliux
- Co-developed an agentic AI chatbot inside Agiliux, an AI-native insurance platform spanning commercial and personal insurance workflows.
- Built RAG and knowledge-retrieval pipelines grounded in client and insurance-domain data.
- Contributed to document-ingestion and OCR workflows that extracted structured information from PDFs and insurance documents.
- Connected internal knowledge and customer data to the conversational layer while maintaining relevant context, permissions, and structured outputs.
- Debugged retrieval quality, document extraction, context handling, and response accuracy where grounding and reliability mattered more than fluent generation.
Builder Mindset
I want to understand why it works.
I am ambitious about building useful systems and eventually starting a company of my own. That goal has made me as interested in the path to adoption as I am in the architecture behind the product.
01 / CURIOSITY
Take the system apart
I like tracing APIs, data flows, and failure modes until I understand why something works, why it breaks, and which constraint actually matters. The task is finished only when the system makes sense.
02 / OWNERSHIP
Build beyond the happy path
I follow a problem through requirements, data, architecture, evaluation, deployment, and real-world edge cases. I test, break, rebuild, and document until the result is useful and dependable.
03 / FOUNDER RANGE
Wear the hat the problem needs
Working toward my own startup has pushed me across engineering, product, marketing, sales, outreach, and distribution. It helps me connect technical decisions to customer value and adoption.
04 / AMBITION
Turn learning into momentum
I am energized by difficult systems, fast feedback, and measurable progress. I learn new domains quickly, communicate across functions, and keep raising the standard once the first version works.
Evidence over buzzwords.
Enterprise outcomes first, followed by selected systems and research with live demos or detailed public case studies.
Selected Evidence
Work that connects data,
models, and users.
Enterprise impact comes first. Independent builds and research follow, with public case studies and live product surfaces where available.
Client Project · Forex Academy
Vault FX Academy
vaultfxtrading.com
Institutional-grade forex education platform built for a client. A structured 12-module mentorship program with a "pay only after you profit" model—live weekly mentor sessions, personal trade reviews, and a private student community. Broker-agnostic, with an animated live trading desk and application flow.
React
Tailwind CSS
Framer Motion
Netlify
Responsive
Live trading desk visualization with real-time EUR/USD positions
12-module curriculum with mentor profiles and trade reviews
Application flow with Tally forms and WhatsApp community integration
Conversion-focused landing with a risk-first, pay-after-profit offer
SaaS Product
Trigrr
trgrr.netlify.app
Closed-loop AI sales coach. Analyzes rep calls against playbooks, detects coaching moments from vocal signals, and schedules AI practice calls with simulated prospects. Full pipeline from call ingestion to Slack-delivered coaching.
Next.js
TypeScript
PostgreSQL
Redis
BullMQ
Deepgram
Vapi
Call scoring with hesitation, pace, and filler word detection
HubSpot, Fathom, Deepgram, Slack integrations
Monday briefing dashboard with pipeline and leaderboard
AI practice calls simulating real prospect objections
AI Platform
HealthFire
healthfire.vercel.app
A deterministic Marketplace subsidy scenario simulator that connects a guided product experience to a Python/FastAPI calculation engine, versioned regulatory data, saved scenarios, and plain-language results.
Next.js 14
Python
FastAPI
PostgreSQL
Regression Tests
Deterministic, explainable calculation path with validated request and response models
Versioned regulatory reference data and golden regression cases
Authentication, row-level access patterns, rate limiting, and audit/security logging foundations
AI Plugin · Premiere Pro
Agentic Premiere
Advanced prototype · Public engineering case study
AI-assisted editing inside Adobe Premiere Pro, designed around a hard trust boundary: the model proposes a typed edit plan, the editor approves it, and the plugin applies and verifies only that bounded plan.
Gemini Vision
UXP SDK
FastAPI
FFmpeg
MCP
UXP panel, Python/FastAPI services, FFmpeg, Gemini Vision, and MCP-backed tool integrations
Token budget model with context caching—under $30/month
2,101 automated tests across the plugin, contracts, middleware, and host-specific workflows
Machine Learning Research · Graph Learning
SG-GNN
Spike-gated graph neural network research
A ten-stage experimental program testing whether Leaky Integrate-and-Fire neurons can learn hard binary edge decisions before graph message passing. The work compares spike gates with sigmoid attention, NeuralSparse, Hard Concrete, and true chemical bonds across QM9 and ZINC.
PyTorch
Graph Neural Networks
Neuromorphic ML
QM9
ZINC
Spike gate reached a 1.3% MAE gap to soft attention while selecting 73% fewer edges in the QM9 ablation
Within 0.3% of NeuralSparse on the controlled ZINC comparison while selecting fewer edges
Reproducible baselines, ablations, fixed-seed runs, comparison tables, and preserved negative results
B2B Marketplace
ScoutPanels
scoutpanels.com
Product validation marketplace connecting founders with working PMs and engineers from Visa, Adobe, and Microsoft. Structured feedback on adoption blockers with video walkthroughs of real usage.
Full-Stack
ML Matching
Lean Startup
Role, domain, company-context, and product-stage matching model
50+ customer discovery interviews, 15+ validators onboarded
Structured friction and time-to-value metrics
Enterprise AI · IndiGo
AskPrism
Internal enterprise product · IndiGo (InterGlobe Aviation)
Enterprise GenAI chatbot serving 200+ daily queries across 5+ departments. RAG-powered retrieval over aviation knowledge bases (CDP, AIMS, HR, Navitaire) with hallucination-reduced outputs for mission-critical airline data. Reduced data query resolution time by 60%.
RAG
LangChain
GPT-4
Azure
Snowflake
Multi-stage retrieval with vector embeddings and semantic search
AI agent workflows with tool-use automating cross-system lookups
RBAC governance for AI-accessible knowledge stores
Enterprise Data Engineering · IndiGo
Flight Operations Data Platform
Internal production platform · IndiGo
Owned the Flight Operations migration from SQL Server to Snowflake and translated operational requirements into procedures, functions, pipelines, and Dynamic Tables. Connected AIMS, crew and roster, Flight Operations, and Navitaire sales data for delay analysis, prediction, forecasting, and pricing POCs.
Python
PySpark
SQL
Snowflake
Azure Data Factory
Power BI
Reduced processing overhead by 40% across enterprise data pipelines
Production patterns using Streams, Dynamic Tables, materialized views, Time Travel, and Jenkins deployments
Source-to-target validation, incident investigation, RBAC, grants, revocations, and controlled PII access
AI Tool
ContextLayer
contextlayerio.netlify.app
AI prompt builder for consulting teams. Input engagement details—client industry, project phase, team size—and get five customized prompts grounded in actual project context. Works with Claude, ChatGPT, Copilot.
LLM Integration
Prompt Engineering
PDF/Markdown
Document upload for SOW, project docs, meeting notes
Multi-LLM support—Claude, ChatGPT, Copilot
No signup required, free to use
Technical Skills
The toolkit
✶
Data Engineering
Python, SQL, PySpark, Snowflake, Azure Data Factory, ETL/ELT, DBT, Dynamic Tables, Streams, Time Travel, Data Warehousing
⚙
Data Science & ML
Python, Pandas, NumPy, Scikit-learn, PyTorch, TensorFlow, Hugging Face Transformers, NLP, Deep Learning, GNNs, Prediction, Forecasting, Experiment Analysis
➤
Applied AI & GenAI
LLMs, RAG, LangGraph, LangChain, Claude Agent SDK, OpenAI APIs, Assistant APIs, FastMCP, MCP, Tool Orchestration, Knowledge Grounding, Context Engineering, Human-in-the-Loop Automation
☁
AI Evaluation, Observability & Governance
MLflow, Braintrust, Langfuse, Tracing, Benchmarking, ROUGE, BLEU, METEOR, Error Analysis, Accuracy/Cost Trade-offs, RBAC, PII Controls, Human Approval, Auditability
Education
B.Tech, Computer Science & Engineering
Maharaja Agrasen Institute of Technology, GGSIPU
2020–2024
Published: Medical Plant Classification Using Deep Learning—IJERCSE, Feb 2024
Looking for a data or AI engineer?
I'm open to full-time Data Science, Applied AI, and AI/ML engineering roles where reliable systems and measurable outcomes matter.