Leadership · Team lead
Hardik Devmurari
AI Agent & ML Engineer — AI Agents, MCP & Machine Learning
Builds AI agents that reason over context, use tools, remember what matters, and hand off to a human when needed — choosing the agent pattern from the business problem. On the ML side, picks and evaluates models, fine-tunes when prompt, RAG, and tools aren't enough, and tests agents against golden datasets before they ship.
Best forAI agents and multi-agent systems, MCP tool integrations, agent memory, and ML work: model evaluation, fine-tuning, and open-source models.
Expertise
What Hardik works in day to day.
AI agents
- LangGraph
- Tool calling
- Multi-agent systems
- Human-in-the-loop
- Agent memory
MCP & tools
- MCP servers & clients
- CRM & calendar tools
- Internal API tools
- Tool permissions
Machine learning
- Model selection
- Fine-tuning (LoRA / QLoRA)
- Open-source models
- Embeddings
- Ollama / vLLM
Evaluation
- Golden datasets
- Tool-call evaluation
- Hallucination checks
- Regression testing
- Model comparison
Typical engagements
The projects Hardik leads for our clients.
- 01AI employees
- 02AI operations agents
- 03Multi-agent platforms
- 04MCP servers for internal tools
- 05Research agents
- 06Model fine-tuning
- 07Agent evaluation suites
- 08Self-hosted model setups
Leadership
One team, six specialisms.
Every project gets the right lead for each part — solution architecture, agents and ML, automation and data, security and cloud, product and engineering, and delivery.
Shyamsinh ParmarAI Solutions & Systems Architect