Hello! I'm

HUZAIFA
ATHAR

Huzaifa Athar - AI / Machine Learning Engineer
Open to opportunities
What I do
AI/MachineLearningEngineer

Ask my AI anything about me

Scroll
About Me

Who I am

What I Do

AI/ML Engineer with 5+ years of software engineering experience, including hands-on work designing, building and deploying production AI systems.

I specialise in LLM applications, multi-agent architectures, RAG and document intelligence, and I own the full pipeline from data preparation and evaluation through containerised cloud deployment and monitoring. I work directly with enterprise clients in regulated industries and can explain model behaviour and progress to non-technical stakeholders.

Track Record
5+Years in engineering
10+Products shipped
15Markets monitored
AI Products

AI & Agents

Agentic compliance pipelines, LLM chatbot platforms with RAG, document intelligence, and voice agents, from design to production.

Leadership

Team & Client Lead

Led cross-functional teams, managed client relationships, drove project planning through execution to delivery. Individual, team, and multi-team scale.

Currently at

Salestech Data & AI

Senior Software Engineer (AI)

Forward-deployed AI engineer embedded with enterprise clients, delivering agentic and data platforms from design to production.

Have a vision?

I'll plan it, build it, and ship it.

Selected Work

What I've Built

AI / Agents

Agentic Compliance Engine

Supervisor-worker multi-agent system for a safety-critical rail systems manufacturer, with every conclusion backed by source evidence and engineer review.

Weeks to minutes
Multi-Agent · Full Traceability
Read case study
AI / RAG

Data Intelligence Chatbot

Engineered a custom RAG pipeline sourcing from a data lake, refining data through transformation layers, and powering an LLM-driven chatbot. Clients query their own knowledge base conversationally and get accurate, context-aware responses in seconds.

Data lake to insight
Custom RAG · LLM Powered
Voice AI

Dental Receptionist AI Agent

Built a Dutch and English AI voice receptionist using ElevenLabs for a dental clinic. Handles incoming calls autonomously, greeting patients, answering FAQs about services, hours, and insurance, routing calls, and booking appointments 24/7.

24/7
Autonomous Call Handling
Voice AI

Appointment Booking Voice Agent

Developed a conversational AI agent that schedules, reschedules, and cancels appointments entirely through voice. Integrates with calendar systems, handles timezone logic, sends confirmations, and manages the full booking lifecycle hands-free.

Voice-first
Scheduling · Calendar · Confirmations
Enterprise Data

Data Quality Guardian: 15 Countries

Built an automated data quality monitoring platform spanning 15 countries. Detects anomalies, generates severity-based alerts, creates tickets automatically, and routes them to the right person based on issue type and severity level.

15 countries
Real-time Monitoring · Automated Alerting
Electric Vehicles

Battery Testing & Certification Platform

Developed an end-to-end platform connecting battery testing labs with EV manufacturers. Labs schedule tests, manage workflows, and deliver certified results, giving both sides a single source of truth from prototype to production-ready battery.

Lab to production
Full Testing Lifecycle
Electric Vehicles

EV Charging Station Locator

Built a mobile application helping electric vehicle owners find the nearest available charging stations in real-time. Features include route optimization, live availability status, and turn-by-turn navigation to the closest compatible station.

Real-time
Station Discovery · Route Optimized
Food & Hospitality

Multi-Branch Restaurant Operations Suite

Built a SaaS platform managing restaurant inventory, custom and general recipes, and supplier coordination across multiple branches. Both restaurant operators and suppliers manage their side of the workflow in one unified system.

Multi-branch
Inventory · Recipes · Suppliers
Case Studies

From problem to production

From problem to production: how I take AI systems from diagnosis to measurable outcomes.

Rail & Manufacturing

Months of compliance reading, done in minutes

Safety-critical rail systems manufacturer, Germany

The Problem

Senior engineers cross-referenced product specifications against regulation and past certifications by hand, taking weeks to months per pass.

What Shipped
  • Agentic pipeline reading specification, regulation and certification history together
  • Supervisor-worker agent architecture with each agent scoped to one job
  • Every conclusion tied to its source documents
  • Engineer review and approval on every output
Outcomes
Weeks to minutesPer compliance cross-reference
100%Conclusions linked to source evidence
100%Outputs reviewed by an engineer
FinTech

Real-time risk scoring and fraud detection

Payments platform, 40M+ transactions per month, US and EU

The Problem

Fraud losses grew faster than volume, models refreshed only quarterly, and manual review took six hours per case.

What Shipped
  • Real-time scoring service (sub-50ms at p99)
  • Graph-based fraud feature store across accounts, devices and merchants
  • Eval harness with golden fraud sets gating every release
  • Analyst case console where the model abstains and humans decide
Outcomes
-42%False positives
11 minCase review, was 6 hrs
DailyModel refresh, was quarterly
Insurance

AI document intelligence and workflow agents

Insurance operations, 60,000 documents per month, 4 systems of record

The Problem

Claims and correspondence were keyed in by hand, with 14-day backlogs and errors caught weeks later.

What Shipped
  • LLM extraction pipeline with confidence-gated outputs
  • Workflow agents filing into four systems of record
  • Exception review console
  • Continuous evaluation against human corrections
Outcomes
4 hrsProcessing, was 14 days
78%Straight-through, from 12%
1,200Analyst hrs/month reclaimed
Healthcare

HIPAA-compliant patient analytics and AI pipelines

Provider network, 12 sites, United States

The Problem

Leadership needed population-health insight, but patient data could not leave the compliance boundary and analyses waited weeks.

What Shipped
  • Governed lakehouse inside the client's own cloud tenancy
  • De-identification at ingestion
  • Readmission-risk models with clinician-readable audit trails
  • Role-based, row-level access
Outcomes
Same dayTime to insight, was 3 wks
0.84Readmission AUC, from 0.68
ZeroPHI incidents in 24 months
Supply Chain

Real-time route optimisation and demand forecasting

Regional 3PL, 400+ vehicles, 6 distribution centres

The Problem

Routes were planned overnight on stale demand data, causing half-empty return trips and stockouts.

What Shipped
  • Live demand model on streaming telemetry
  • Event-triggered route re-optimisation
  • SKU-level forecasts wired into replenishment
  • Planner console with auditable overrides
Outcomes
-18%Cost per route
93%Forecast accuracy, from 71%
-37%Stockout events, YoY
Experience

Where I've worked

Salestech Data & AI

NOW

Senior Software Engineer (AI)

Dec 2025 - Present

Built an agentic compliance pipeline for a safety-critical rail systems manufacturer, using a supervisor-worker multi-agent architecture (GPT-4o, Claude). It cuts certification cross-referencing from weeks to minutes, with every conclusion traced back to its source documents and an engineer approval step. Also built a data quality platform for a pharmaceutical client monitoring 15 European markets, with automated alerting.

PythonLangGraphGPT-4oClaudeMulti-AgentRAGAzure

DigitLabs

Full Stack Engineer (AI Products)

Sep 2024 - Dec 2025

Built the Django backend and API layer for an LLM-powered chatbot platform, integrated ML-driven insights with data scientists, built analytics dashboards, and delivered a battery-testing platform connecting OEMs and testing labs.

DjangoPostgreSQLCeleryRedisLLMs

MTP

Software Engineer

Feb 2022 - Aug 2024

Built full-stack applications across multiple products: Django & FastAPI backends, React & Next.js frontends. Created high-performance microservices and managed AWS infrastructure.

DjangoFastAPIReactNext.jsAWS

Codegic

Associate Software Engineer

Aug 2020 - Feb 2022

Led frontend development with React, built reusable component libraries, and established structured backend query patterns. Active in Agile ceremonies and code review culture.

ReactJavaScriptDjangoPostgreSQL
AI Voice Agents

Voice AI Products

I build intelligent voice agents that handle real conversations, from dental reception to appointment scheduling.

Voice Recognition
NLP Processing

Dental Receptionist AI

An AI-powered voice agent that handles incoming calls for dental clinics, greeting patients, answering FAQs about services, hours, and insurance, and routing calls to the right department.

Natural conversation flowFAQ handlingCall routing24/7 availability

Appointment Booking Agent

A conversational AI agent that schedules, reschedules, and cancels appointments via voice. Integrates with calendar systems, handles time zone logic, and sends confirmations.

Smart schedulingCalendar integrationRescheduling & cancellationConfirmation notifications
Tech Stack

My Skills

Contact

Let's work together

Got a project in mind? I'm always open to discussing new opportunities and ideas.