This post is part of the Building It series where I share my journey building an industrial-tech startup, Amygda. Published weekly and as transparent as our customers, investors, and partners would have us be (read: legally allowed).
We are now in the period of accelerating product development. Building is unlike other things in a startup. Building (the product) requires focus and is at the core of solving deeply technical challenges.
So I won’t say this is a 3 month or a 6 month affair. To be honest, we are perpetually building – that’s the whole point of being agile and CI/CD.
But the next 6 months are going to have product heavy focus.
This final quarter of 2020 is definitely product focussed. And a few other things. If you want to know our objectives of the final quarter of 2020 read this blog.
Deep tech product development is hard
At Amygda, we are at the start of the product development journey. Building from scratch is hard. It’s time consuming, but also as a startup, in the B2B space you are fighting two battles:
- Building your own product with a vision that you have, solving challenges that come up with trying something new
- And building it in a manner that doesn’t completely change the human interaction (as re-training employees is a cost).
We are definitely trying to balance the two points. Because the maintenance and service engineers are be users of our product don’t like change. There’s no point in us trying to fight that. It’s acceptance.
And yet, we are changing how things are done. But keeping that behind the scenes.
So our approach is, the whole back-end system and how things are done is where we innovate. And then use the front-end to give our users the warm feeling that things are the same, and they can hop onto this product.
Most of the time, we are getting our users away from tools like
- Excel (freaking hell this is a beast of a tool to migrate them away from)
- Emails (people manually do repetitive tasks and notifications)
- Or some BI tool, completely lacking in depth but looked good as innovation thrusted upon them
And then some of the other challenges to solve
Which repeatedly you will encounter in B2B space is around data ingestion (or acquisition).
Amygda’s platform makes data-driven decisions. It means we do rely on data pipelines from existing infrastructure.
Before we start processing and number-crunching, we expect two things to be in place
- The customer already acquires sensor data
- Customer uses some ERP system (SAP, ORACLE, Cloud) or a historian DB (like Excel even) that we can ingest data from
And each customer can have slightly different data formats. As we’ve discovered, data formats aren’t that different when you are working in the same vertical.
Most verticals have 3/4 OEMS and once we work with equipment across them, we will have a library of data connectors in place to reduce the work in connecting to different data types. But it doesn’t exist now.
A note to investors or founders: A lot of investors ask us about data acquisition and how we will reduce that work, to make the product scalable. There are only two ways to do this. First is to either build and/or deploy the sensors (data acquisition modules) yourself. So you can capture data in a specific manner. Or secondly, you have to put the effort in to connect to a broad group of acquisition systems. There isn’t a third way. We decided to focus on the second solution, because building your own sensors (hardware) is not something we want to do. And the best investors in enterprise space know this. Look at Uptake, Palantir, or speak to any subject matter expert in building data platforms and they will say the same thing.
The summary of this part is that we are heavily focussing on product development. And there are some long days and months ahead of us (this bit is added for dramatic effect, nothing more).
So how’s the recruiting for the full-stack engineer coming along?
It’s coming along well (work in progress). Thank-fully we weren’t inundated with applications. To be honest, at one point I was scared if anyone would even bother wanting to join us.
I was surprised when some folks did apply 🙂
Oh, and the job description we put out went through a number of changes as we went along. I did learn a few things along the way. Some of the learnings were:
- Recruiting is a full time job if you want to scout for candidates
- Good recruiters do add value, we didn’t use any but did speak to Alan Furley, who offered his advise and helped me improve the job description and our recruitment approach (thanks Alan, if you read this :D)
- Recruiting for a role takes time, this is partly driven by being early-stage, (ahem salary discussion) and by not focussing on it enough before this point (something I’ve learnt from)
- And damn, is it expensive. I know it is not expensive in the grand scheme of things falana falana, but hey, when you are trying to build a business in the middle of a pandemic, and a recession in the economy, and being pre-revenue, and in pre-seed, you are living frugally and everything is expensive. The answer is, budget well for it as you plan everything else.
So what’s the update on the role, did you find anyone?
We are having some great discussions with a couple of technical people and we should have the selection done soon.
On the role itself here are some learnings in our search for the full-stack engineer
- Zero women applied for the role, even though I actively encouraged everyone to apply
- A few engineers I spoke to wanted to be second engineer on the team, but not ready for lead position
- At our stage we are 12 – 18 months of long days and heavy commitment, this wasn’t for everyone depending at the stage of career they were at
- No one wants to join us for the salary, all of the candidates were making more, but they want to be part of the adventure
- Empathy was a common thread in everyone we spoke to, empathy is very important for us. Douchebags are never welcome at Amygda
- Remote hiring is very different to in-person and there is always a bit of risk, but we live on risks right? (this is another dramatic statement, but hey I am the author)
My aim was to get to final list of candidates by the end of October and have the agreement in place before end of November. And we are on track.
More to follow on this as we progress.
But my search to have a woman join Amygda full-time continues. And I want this because otherwise I would be a hypocrite asking companies to diversify.
I know about all the theories about “best person for the job” etc etc. But men can’t always be the “best person” for the job can they? There’s inherently a mistake in the / our(?) system if we fail to have some balance.
I was hoping to get back to Papatya (a good friend, and gives good feedback, but really, she is just very smart, so I listen) and say how I’ve dunked on her feedback (that I need to still work on diversity aspect), and we are now a diverse founding team. But nope, she still has one point over me.
To summarise up to this point
Product development is hard. There is a challenge of product build itself. But also the challenge of attracting a diverse and right skillset of people to take on the challenge.
Any update on the customers?
This is already a long post, and I don’t want to bore you more.
But there is no major update on customers from the previous update at week 14.
So what’s happening next?
Couple of interesting things ongoing that I am experimenting with, one of them is getting super-organised and hyper-disciplined.
Trust me, it feels boring to be that way. I don’t even know if there is life outside some days.
But we have so much ongoing, and we could do with couple extra hands helping us. However, that’s not a luxury (read: money) we have right now.
On discipline, it’s actually not as draconian as it sounds. But there are some things I am trying out, which I will share with you next time around on how they go.
…
What do you think of this week? Let me know by messaging me on LinkedIn or tweet me @faizanpatankar. I love talking, just mention you read my blog.
Related blogs
A month of clusterf*ck behind the scenes, but raised funding
Why do I openly share all our moves about building an industrial analytics startup?