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 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.