Avahi

BACKGROUND

Please describe your company and position.

I am the CEO of Skulpty

Describe what your company does in a single sentence.

We make Skulpty, an AI tool that generates production-ready 3D models from a simple prompt, so people can skip the slowest and most expensive part of building in 3D.

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Avahi to accomplish?

• Assist in building the machine learning pipeline behind Skulpty so we could turn prompts into game-ready 3D models.
• Set up the preprocessing and multi-view rendering workflow to get our training data ready at scale.
• Bring real ML expertise onto the project to help us fine-tune the models and move faster than we could alone.

SOLUTION

How did you find Avahi?

Referral

Why did you select Avahi over others?

• High ratings
• Good value for cost
• Referred to me
• Company values aligned

How many teammates from Avahi were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

Avahi worked with us on the machine learning side of Skulpty, our platform that turns text and image prompts into game-ready 3D models. The core of the engagement was building out the pipeline that makes that possible, from preparing our training data to fine-tuning the models that generate the actual 3D output.

A big part of the work was the preprocessing side. They set up a workflow that took our source assets and generated the multi-view renders we needed, then stored and organized everything so it was ready to train on at scale. That groundwork mattered because the quality of what comes out of a model like this depends heavily on how clean and consistent the inputs are going in.

From there they helped us fine-tune the model itself, working through the ShapeVAE and DiT stages and iterating with us as we tested results. Their ML lead was hands-on throughout and brought a level of depth we didn’t have in-house, which kept us moving instead of getting stuck. The main deliverables were a working preprocessing and rendering pipeline, a fine-tuned model producing usable 3D assets, and the technical guidance to keep improving it after their part was done.

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

The milestones so far have been real. The biggest one is that we now have a more polished machine learning pipeline where the previous pipeline needed refinement, which is the foundation everything else depends on. Along the way, the preprocessing and multi-view rendering workflow has let us prepare training data far faster than handling it manually, and the fine-tuning work is already producing 3D output we can build the rest of the product around. Having their ML lead embedded with us has kept us moving and hitting our technical milestones instead of stalling on problems we couldn’t solve alone. The clearest sign of progress is simple, the core of Skulpty is coming together and working, and the collaboration is still active as we push toward the full MVP.

Describe their project management. Did they deliver items on time? How did they respond to your needs?

Since we’re still building the MVP together, project management has been an ongoing, collaborative rhythm rather than a one-and-done handoff, and that’s been one of the strengths. They’ve kept the work moving at a steady pace and have been good about hitting the technical milestones we set along the way. Communication has been the standout. Their ML lead has stayed embedded with us, so when something comes up we’re not waiting days for a response, we work through it together in real time. They’ve been responsive to our needs and quick to adjust when we’ve changed direction or run into a problem, which matters a lot when you’re building something this technical from the ground up. It’s felt less like hiring a vendor and more like having an extension of our own team.

What was your primary form of communication with Avahi?

• Virtual Meeting
• Email or Messaging App

What did you find most impressive or unique about this company?

What stood out most was the depth of real machine learning expertise they brought to the table. A lot of firms can talk about AI, but Avahi actually understood the hard parts of building a generative 3D pipeline, the data preprocessing, the model fine-tuning, the tradeoffs that only show up once you’re deep in it. That level of specialization is rare, and it’s exactly what we needed. The other thing that set them apart was how they worked with us rather than just for us. Their ML lead embedded with our team and treated the problem like it was his own, which made the whole thing feel like a true partnership instead of an outsourced task. For an early-stage company building something technically ambitious, having a partner who’s

Their commitment to the mission, without ever sacrificing quality or the things that matter most when building a product, has been refreshing, and I’d highly recommend them to anyone who needs the same.

Are there any areas for improvement or something Avahi could have done differently?

No complaints.

BACKGROUND

Please describe your company and position.

 I am the CEO of an information technology company 

Describe what your company does in a single sentence.

Business Intelligence for Real Estate Brokers

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Avahi to accomplish?

• Propensity scoring

SOLUTION

How did you find Avahi?

Referral

Why did you select Avahi over others?

• Pricing fit our budget
• Referred to me

How many teammates from Avahi were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

The engagement focused on the development of a GenAI-powered proof of concept involving data enrichment, scoring workflows, API integration, and cloud-based AI infrastructure. The scope included the design and implementation of backend scoring pipelines, integration with external data sources, AI-powered explainability workflows, and a functional frontend/dashboard interface. Key deliverables included a working scoring and enrichment pipeline, API endpoints, AI explanation workflows, supporting documentation, and technical handover materials to support future internal development and expansion.

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

The project successfully delivered a functional GenAI proof of concept within the planned engagement period, including working data enrichment pipelines, deterministic scoring workflows, API-based outputs, and AI-powered explanation capabilities. Additional outcomes included the establishment of a reusable technical foundation for future AI initiatives, improved internal understanding of scoring and enrichment workflows, and the successful integration of multiple external data sources into a unified processing pipeline.

Describe their project management. Did they deliver items on time? How did they respond to your needs?

The Avahi team was responsive, collaborative, and supportive throughout the engagement. They communicated regularly with our team, adapted to evolving technical requirements, and were proactive in working through implementation and architecture discussions. Overall, the team demonstrated strong technical execution and maintained a professional and solutions-oriented approach throughout the project. While some handover and transition items required additional coordination toward the end of the engagement, the team remained responsive and supportive in helping us work toward project closure.

What was your primary form of communication with Avahi?

Virtual Meeting

What did you find most impressive or unique about this company?

What stood out most was the team’s speed of delivery and willingness to iterate quickly. They were able to move from initial discussions to a working GenAI proof of concept within a relatively short timeframe, while remaining collaborative and responsive throughout the process.

Are there any areas for improvement or something Avahi could have done differently?

One area that could be improved is the transition and handover process toward the end of the engagement. Additional alignment earlier around final documentation, repository handoff, and project closure expectations would have helped streamline the final stages of the project.

BACKGROUND

Please describe your company and position.

I am the CTO of myRiva

Describe what your company does in a single sentence.

myRiva is a leading corporate travel marketplace, and it deliberately broke from the traditional model. Instead of acting as a travel management company, myRiva runs as a marketplace, letting travelers book directly with their preferred hotel or airline. That direct-booking approach means travelers keep the loyalty points they earn, with a seamless experience across every source. At the same time the company keeps full control: myRiva still manages the program and enforces travel policy on the organization’s behalf. Travelers get what they want, and the business keeps its guardrails.

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Avahi to accomplish?

• Automate QA
• Automate Code Review

SOLUTION

How did you find Avahi?

Referral

Why did you select Avahi over others?

• High ratings
• Other

How many teammates from Avahi were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

Phase 1: Discovery & Architecture Definition (Week 1) Objectives • Align on MyRiva’s QA goals, development workflows, and risk areas. • Define an AI-assisted architecture that integrates naturally into existing tools. In-Scope Activities • Project kickoff and requirement walkthrough. • Review of: ◦ Application architecture (frontend, backend, middleware). ◦ Existing QA processes and historical failures. ◦ CI/CD pipeline (Bitbucket + Jenkins). ◦ Observability stack (Grafana, Graylog, New Relic). • Architecture design covering: ◦ AI-generated test case strategy (code-aware, change-driven). ◦ Use of AI coding assistants for: ▪ Bug analysis ▪ Unit test generation ▪ Fix suggestions ◦ Human-review workflows for AI recommendations. ◦ Observability data ingestion for defect detection. • Definition of PoC success metrics. Deliverables • Architecture and solution design document. • Definition of PoC success metrics. 
Phase 2: AI-Assisted QA & Coding Automation (Week 2) Objectives • Implement core AI-assisted workflows focused on QA, bug fixes, and PR reviews. In-Scope Activities • AI-Generated QA Automation Page 5 of 14 Document Ref: 5SSZK-CDGMB-BDLBA-QTOHF Page 5 of 14 Project: myRiva - Avahi - GenAI POC ◦ AI-driven test generation based on: ▪ Code structure and parsing. ▪ Code comments and annotations. ▪ Recent code changes and diffs. ◦ Initial focus on 50% automated QA coverage for selected services. ◦ Tests generated and validated via developer review (no blind automation). • Coding Assistant for Bug Fixes & Support Tickets ◦ Use of AI coding assistants and agent-based tools to: ▪ Analyze support tickets and related code paths. ▪ Propose bug fixes and refactoring suggestions. ▪ Generate unit and regression tests to validate fixes. ◦ Examples of tooling approaches: ▪ IDE-embedded copilots (Cursor-like). ▪ Task-oriented coding agents (Devin-style). ▪ Multi-step reasoning agents for large codebases (Kiro / Antigravity-like). ◦ Goal: reduce time-to-fix and human error. • AI-Assisted Code Review ◦ Automated PR analysis to: ▪ Detect potential vulnerabilities. ▪ Flag risky changes or anti-patterns. ▪ Suggest remediations and best practices. ◦ Integration aligned with Bitbucket and Jenkins workflows. ◦ All recommendations remain advisory and review-based. Deliverables • Demonstrated AI-generated tests. • AI-assisted bug fix and unit test examples. • Automated code review insights integrated into PR flow. 
Phase 3: Observability Integration, Validation & Documentation (Week 3) Objectives • Automated code review insights integrated into PR flow. • Validate PoC outcomes and document learnings. In-Scope Activities • Integrate observability and tracing platforms (Grafana, Graylog, New Relic) to: ◦ Identify recurring errors and anomalies. ◦ Highlight performance and scaling risks. • Use AI to: ◦ Summarize log patterns. Page 6 of 14 Document Ref: 5SSZK-CDGMB-BDLBA-QTOHF Page 6 of 14 Project: myRiva - Avahi - GenAI POC ◦ Suggest potential root causes. ◦ Recommend areas for test coverage or refactoring. • End-to-end validation of: ◦ QA automation ◦ Coding assistance ◦ Observability-driven insights • Documentation and knowledge transfer. Deliverables • Observability-driven issue insights. • Final PoC documentation and handover

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

AI generated tests, AI-assisted bug fix and unit tests, automated code review integrated into PR flow, ened-to-end validation of QA austomation.

Describe their project management. Did they deliver items on time? How did they respond to your needs?

They were always professional, demonstrated a great understanding of our goals and delivered what they promised when they promised it.

What was your primary form of communication with Avahi?

Virtual Meeting

What did you find most impressive or unique about this company?

They didn't treat this as just a set project.  They understand that it served as a gateway into a larger technology initative of myRiva and ensured their efforts not only supported the project but the overall roadmap.

Are there any areas for improvement or something Avahi could have done differently?

None

BACKGROUND

Please describe your company and position.

 I am the CEO of a CRM company 

Describe what your company does in a single sentence.

AI-native CRM for government teams

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Avahi to accomplish?

• Business card OCR scan to CRM
• Social media comments pull to CRM

SOLUTION

How did you find Avahi?

Referral

Why did you select Avahi over others?

• High ratings
• Pricing fit our budget
• Good value for cost
• Referred to me

How many teammates from Avahi were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

Delivered code and documentation for two projects: (a) business card scan and (b) social media comments pull

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

Delivered the code and documentation; not implemented in prod yet

Describe their project management. Did they deliver items on time? How did they respond to your needs?

Delivered on time; very responsive

What was your primary form of communication with Avahi?

• Virtual Meeting
• Email or Messaging App

What did you find most impressive or unique about this company?

They understood our project goals quickly and developed it with minimal input.

Are there any areas for improvement or something Avahi could have done differently?

None

BACKGROUND

Please describe your company and position.

I am the Founder of an HR coaching company.

Describe what your company does in a single sentence.

We're a leadership coaching and workforce development training company focused on helping organizations strengthen belonging, resilience, and performance—while giving leaders and teams the tools they need to thrive in today’s workplace.

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Avahi to accomplish?

• Ai Agent Development
• AWS Integration

SOLUTION

How did you find Avahi?

• Referral
• Other

Why did you select Avahi over others?

• Referred to me
• AWS Partner Recommendation/Program

How many teammates from Avahi were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

Three completely different client environments, a government accelerator program, a public utility, and a global tech company, each with a different agent persona, tone, and coaching focus, all running on the same core pipeline.

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

We came to Avahi with a complex ask: a multi-tenant AI coaching platform where each client's agent would have a distinct personality, knowledge base, and coaching focus, all dynamically configured at login based on the company's domain. Finding a technical partner who could understand the human-centered mission behind the platform AND execute it with serious AWS infrastructure expertise is rare. Avahi delivered both.

Within a three-week engagement, they architected and built a working proof-of-concept using Amazon Bedrock and Anthropic Claude 3 Sonnet, with a RAG pipeline built on Amazon OpenSearch that allows each tenant's AI agent to pull from its own distinct knowledge base.

Describe their project management. Did they deliver items on time? How did they respond to your needs?

On time, above expectations, and also was accomodating for additional modification requests. 

What was your primary form of communication with Avahi?

Virtual Meeting

What did you find most impressive or unique about this company?

Avahi understood my company's needs, goals, and objectives, had prior experience with our request, made suggestions, and even translated desires into optimized project elements. They were highly recommended through an Ai-Tech Founder community and I was looking specifically for an AWS partner.

Are there any areas for improvement or something Avahi could have done differently?q

For founders and early-stage startups, the onboarding and technical setup process can feel fast-moving. A lightweight orientation session at the start of the engagement to walk non-technical clients through the AWS environment setup would be a valuable addition to their process. Avahi's expertise is clear, and bridging that gap for clients who are building their technical infrastructure for the first time would make an already strong experience even smoother.