Aashish Pant
PART NO. AP-LLM-26
Kathmandu, Nepal · UTC+5:45
Kathmandu, Nepal · UTC+5:45
AI engineer building production-grade agentic systems
STATUS: AVAILABLE
1Operating conditions
| Parameter | Value |
|---|---|
| Role | AI engineer |
| Domains | LLMs, AI agents, finance, AI automation across domains |
| Location | Kathmandu, Nepal · UTC+5:45 |
| Mode | Remote or on-site |
| Currently | DevDash Labs |
| Toolchain | OpenAI Agents SDK, LangGraph, LangSmith, promptfoo, PyTorch, XGBoost, CatBoost, FastAPI, Django, PostgreSQL, Docker, Terraform, AWS, GCP |
2Specifications
| Parameter | Conditions | Value |
|---|---|---|
| Research brief cost, reduction | tested workloads, A/B verified | >50% |
| Agent pipeline latency, reduction | sequential → parallel | ~10× |
| Shipping production software | since mid 2023 | 3 yr |
| Fusemachines AI Fellowship | national applicant pool | top 3% |
3Typical applications
| Application | Where | Result |
|---|---|---|
| Identify: research agent that maps a whole market | Alkemy | 600+ companies in a domain found in a single sweep |
| Key account intelligence agent | Alkemy | Pre-meeting briefs, open issues and attendee prep for each account |
| Evaluation suite for a multi-agent platform | Alkemy | Caught a tool reporting success on fetches that were refused |
| Context and cost redesign | Alkemy | Cost per brief more than halved |
| TradePilotAI: parallel agents that propose options trades | DevDash | ~10× faster than the sequential CrewAI version |
| TradePilotAI: multi-model prediction pipeline | DevDash | Per-session XGBoost and CatBoost models, embargoed validation |
| Natural language to MongoDB query generator on T5 | Research | Schema-aware T5, trained on a self-checked synthetic corpus, beat a schema-agnostic baseline on complex operators |
| IoT crop recommendation: ESP8266 sensor network and an SVM | IOE | Live soil and field readings in, a crop recommendation out, served over FastAPI |
| Multi-tenant gift certificate platform | Atrusta | Built end to end, Stripe via webhooks |
| Booking assistant for a skincare clinic | Atrusta | Won't invent services or prices |
4Revision history
| Rev | Date | Change |
|---|---|---|
| D | Apr 2024 – now | DevDash Labs, Member of Technical Staff, Applied LLM Systems |
| C | 2023 – 2024 | Atrusta, Product EngineerPart-time since |
| B | 2023 | Fusemachines AI Fellow, ML/AI research track |
| A | 2019 – 2024 | BE, Electronics, Communication and Information EngineeringIOE Thapathali, Tribhuvan University. First Division. |