You will get AI agent with tool use, memory, and multi-step reasoning in your app
Rising Talent
Rising Talent
Project details
An AI chat widget is a feature. An AI agent is a system that can think through a problem, decide what to do, call your APIs, remember previous interactions, and produce a result without hand-holding at every step. I build production AI agents using Claude's tool use and multi-step reasoning capabilities.
Agents are hard to build correctly. Most break on edge cases: tool calls that return unexpected data, reasoning loops that don't terminate, context windows that overflow, state that gets corrupted between turns. I've designed agentic systems for production apps and know how to build the guardrails that keep them reliable.
What I build:
• Agent reasoning loop (plan, tool call, observe, plan again)
• Tool definitions connected to your real APIs and DB
• Short-term memory (conversation context management)
• Long-term memory (vector search or structured DB storage)
• Agent guardrails (max steps, cost limits, human-in-the-loop checkpoints)
• Streaming agent output so users see progress in real time
• Logging and observability for debugging agent behaviour
Agents are hard to build correctly. Most break on edge cases: tool calls that return unexpected data, reasoning loops that don't terminate, context windows that overflow, state that gets corrupted between turns. I've designed agentic systems for production apps and know how to build the guardrails that keep them reliable.
What I build:
• Agent reasoning loop (plan, tool call, observe, plan again)
• Tool definitions connected to your real APIs and DB
• Short-term memory (conversation context management)
• Long-term memory (vector search or structured DB storage)
• Agent guardrails (max steps, cost limits, human-in-the-loop checkpoints)
• Streaming agent output so users see progress in real time
• Logging and observability for debugging agent behaviour
AI Algorithms
Large Language Model, Multimodal Large Language ModelAI Applications
AI Chatbot, Conversational AI, Natural Language GenerationAI Tools
Azure OpenAIAI Models
ChatGPT, GPT-4, LaMDA, LLaMA, WhisperWhat's included
| Service Tiers |
Starter
$349
|
Standard
$549
|
Advanced
$849
|
|---|---|---|---|
| Delivery Time | 3 days | 5 days | 7 days |
Number of Revisions | 2 | 3 | 3 |
AI Model Integration | - | - | - |
Batch Normalization | - | - | - |
Database Integration | - | - | - |
Detailed Code Comments | - | - | - |
Image Upscaling | - | - | - |
MLOps | - | - | - |
Model Deployment | - | - | - |
Model Documentation | - | - | - |
Model Monitoring | - | - | - |
Model Testing & Optimization | - | - | - |
Model Tuning | - | - | - |
Natural Language Processing | - | - | - |
NLP Tokenization | - | - | - |
Pre-Training | - | - | - |
Prompt Engineering | - | - | - |
Setup File | - | - | - |
Source Code |
20 reviews
(19)
(1)
(0)
(0)
(0)
This project doesn't have any reviews.
RM
Robert M.
Mar 21, 2024
Flutter mobile build and ongoing upgrades
Sailesh is one of the best engineers i have ever worked with. He has deep skills across many software engineering practises. We could leave major technical decision making to him, as well as lean on him to lead the build and launch of multiple products.
PV
Pedro V.
Aug 25, 2023
Flutter Mobile App Debugging for App Store Submission
I've troubleshot my Flutter app for some weeks and there were a few different and unexpected bugs. Sailesh was an amazing support throughout this process because he always knew how to solve these niche issues in a matter of hours. It's been a pleasure working with him and I appreciate the communication and relationship established along the way.
PV
Pedro V.
Aug 23, 2023
Flutter: Location Backend Fix
Always a very positive experience working with Sailesh. Very skilled, attentive to detail and great availability.
PV
Pedro V.
Jul 20, 2023
iOS - UI Unresponsiveness Error
Trust is very important when working with someone from Upwork and I am very happy for the relationship I have with Sailesh. For this project, he provided assistance beyond the scope and assisted in providing alternative solutions. Always a pleasure working together
MS
Matthew S.
Jul 11, 2023
Sailesh was a great! He helped us fix some issues with our Android App. From the moment we started chatting I knew he would be good. He was able to understand instructions clearly and provide critical feedback. He was quick in his work and ongoing communication was excellent. Completed the job 100% successfully. Would highly recommend.
About Sailesh
React Native & Flutter Developer | AI Integration, Mobile + Full-Stack
100%
Job Success
Dhulikhel, Nepal - 6:40 pm local time
I've been the sole mobile engineer on products that went from nothing to real users — handling everything from first commit to App Store. React Native, Flutter, Expo, Vite, Express, Hono, MongoDB, PostgreSQL. I work across the full stack so you're not waiting on five different people to ship one thing.
I use Claude Code and Codex daily. Scaffolding, boilerplate, debugging loops — the stuff that eats hours. Doesn't replace thinking. But I move faster because of it.
WHAT I'VE BUILT
→ OneTake — mobile lead for a year. Every pivot, every weekend bug, handled. The founder flew out to meet me. His words: "the backbone of OneTake."
→ Kalz — AI nutritionist app, React Native + Expo
→ Solas Compliance — full web and backend for a compliance SaaS
→ Sambad, Flexera — shipped and live
→ ReceiptSasa, Herd — shipped, real users, wound down
AI WORK
→ OCR pipelines and document processing
→ OpenAI and Gemini API integration — production use, not demos
→ Workflows that automated the repetitive stuff clients were doing by hand
→ Claude Code, Codex, Cursor — daily use, not occasional
STACK
→ Mobile: React Native, Expo, Flutter, Dart
→ Web: React, Vite, TypeScript
→ Backend: Express, Hono, Node.js
→ Database: MongoDB, PostgreSQL
→ Cloud: Firebase, AWS, GCP
I reply within 12 hours. Scope is clear? I can usually start the same day.
Drop me a message — glad to take a look at what you're building.
Steps for completing your project
After purchasing the project, send requirements so Sailesh can start the project.
Delivery time starts when Sailesh receives requirements from you.
Sailesh works on your project following the steps below.
Revisions may occur after the delivery date.
Build your AI agent with tool use and memory