TW

About

Partnering with teams to ship dependable AI systems

I am an AI consultant and ML engineer who helps founders and product leaders turn LLM ideas into production-grade products. My focus: retrieval pipelines, agentic workflows, voice copilots, and evaluation harnesses that keep quality, latency, and cost under control after launch.

Book a build call

How I work

  • End-to-end builds: architecture, rapid prototyping, evaluation harnesses, and production rollout with observability and safety.
  • Deterministic control planes for agents: typed tool contracts, retries, validation, and replayable traces to keep incidents down.
  • Reliability first: golden test suites, cost/latency budgets, and feature flags on retrievers, rerankers, and models.

25+

Production AI launches

7s → <2s

Latency reduced

95%

Infra cost saved

85% faster

Ops time saved

Tallha Waheed
Sterling, Virginia

What I build

Grounded copilots, agentic automation, and realtime AI that ship with tracing, evals, and deployment playbooks so your team can own them.

RAG and search with rerankers, safety checks, and latency budgets

Agents that call your APIs and tools with validation and replayable traces

Voice and chat experiences tuned for sub-2s latency, with dashboards and SLAs

LLM and AI

  • RAG pipelines, rerankers, deduplication
  • Tool-calling agents and workflow graphs
  • Voice AI, diarization, and synthesis routing
  • LLM evaluation harnesses with safety checks

Languages and frameworks

  • Python
  • TypeScript/JavaScript
  • Go
  • Node.js
  • React/Next.js

Data and infra

  • PostgreSQL, Redis, MongoDB
  • Kafka, event-driven pipelines
  • Docker, Kubernetes, AWS/GCP
  • Feature flags, tracing, and observability

Tools

  • LangChain/LangGraph, LlamaIndex
  • OpenAI, Gemini, Deepgram, Whisper
  • Pydantic/Valibot validation layers
  • CI/CD with golden tests and canaries

Experience

Lead AI Consultant

Self-employed

2025 — Now

Partnering with founders, product leads, and engineering managers to deliver end-to-end AI systems.

  • Hands-on builds for RAG platforms, agentic workflows, and voice copilots
  • Advisory blocks for AI strategy, roadmaps, and staffing plans
  • Embed with teams to ship reliable features with observability and cost guardrails

Senior ML Engineer

Accenture

2022 — 2025

Led ML initiatives from architecture to launch with a focus on reliability and performance.

  • Delivered LLM-powered automation and analytics across regulated domains
  • Mentored engineers on evaluation, safety, and production readiness
  • Drove cost and latency optimizations across deployed AI services

ML Engineer

Dropbox

2019 — 2022

Built data products and ML services to improve collaboration and search experiences.

  • Shipped ranking and recommendations that improved engagement
  • Improved data quality pipelines and experiment velocity
  • Collaborated with product and infra teams on scalable services

Data Scientist

Incedo

2016 — 2019

Developed predictive models and analytics for enterprise customers.

  • Delivered credit risk and churn models with explainability
  • Built dashboards and data pipelines to keep models fresh
  • Partnered with stakeholders to translate business needs into shipped models

Available

Let's ship your next AI release

Email devtallhawaheed@gmail.com or connect on LinkedIn for build work, advisory blocks, or staff-aug engagements.